International Journal of Advanced Engineering, Management and

Science (IJAEMS)
Peer-Reviewed Journal
ISSN: 2454-1311 | Vol-9, Issue-10; Oct, 2023
Journal Home Page: https://ijaems.com/
DOI: https://dx.doi.org/10.22161/ijaems.910.2

Deep Learning Neural Networks in the Cloud

Burhan Humayun Awan

[email protected]

Received: 05 Sep 2023; Received in revised form: 08 Oct 2023; Accepted: 15 Oct 2023; Available online: 25 Oct 2023

Abstract— Deep Neural Networks (DNNs) are currently used in a wide range of critical real-world
applications as machine learning technology. Due to the high number of parameters that make up DNNs,
learning and prediction tasks require millions of floating-point operations (FLOPs). Implementing DNNs
into a cloud computing system with centralized servers and data storage sub-systems equipped with high-
speed and high-performance computing capabilities is a more effective strategy. This research presents an
updated analysis of the most recent DNNs used in cloud computing. It highlights the necessity of cloud
computing while presenting and debating numerous DNN complexity issues related to various architectures.
Additionally, it goes into their intricacies and offers a thorough analysis of several cloud computing
platforms for DNN deployment. Additionally, it examines the DNN applications that are already running on
cloud computing platforms to highlight the advantages of using cloud computing for DNNs. The study
highlights the difficulties associated with implementing DNNs in cloud computing systems and provides
suggestions for improving both current and future deployments.
Keywords— Deep Learning, Neural Networks, Cloud Computing.

I. INTRODUCTION servers powered by cloud computing provide a lot of

Numerous pattern recognition applications in real-world computing power, a lot of data storage, fast processing, low
fields like e-commerce, manufacturing, medicine and latency, and high availability. DNNs for online applications
health, and autonomous vehicles are now being developed can be deployed thanks in large part to cloud computing.
using deep neural networks (DNNs). However, due to their
extensive parameter requirements, DNNs pose significant
computing demands, especially during training. DNNs
typically have millions of parameters. As an illustration,
popular DNNs like AlexNet have 60 million parameters,
while VGG-16 has 138 million. A DNN with 175 billion
parameters that required a lengthy seven months to train
was used in a recent OpenAI project for natural language
processing (NLP) [1]. As a result, it is impossible to train a
large DNN using a single isolated computer. High- Fig.1: Deep Neural Network with hidden layers
performance computing tools are necessary for the efficient
training of DNNs. The implementation of cloud platforms for computationally
DNN deployment on cloud platforms has increased in intensive tasks is discussed in various recent survey
popularity recently. These cloud computing platforms are publications cited in the current manuscript. These survey
extremely fast and memory-capable high-performance papers are divided into three main categories: applications,
computing systems. On a variety of cloud machine learning performance improvement technologies, and security
(ML) platforms, such as Google Colab and Amazon Web technologies. Yan et al.'s [2] study of security technologies
Services (AWS) Deep Learning, training can be effectively covered a range of solutions for preventing harmful
carried out in reasonable amounts of time. Centralized assaults, including talks of countermeasures. Recent

This article can be downloaded from here: www.ijaems.com 9

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

security methods, including encryption, access structure, [11] provided an overview of mobile cloud topologies and
and fine-grained trace mechanisms, were introduced by the advantages of cloud computing. An overview of cloud
Nita and Mihailescu [3] and Sun et al. [4]. Gai et al.'s [5] computing architectures for cyber-physical systems was
analysis of the functionality of blockchain-enabled presented by Cao et al. [13], evaluating numerous
integrated hardware and software in cloud data centers was applications. Notably, the development of DNNs in cloud
centered on the integration of blockchain with clouds. computing systems is not a specific emphasis of these
Regarding performance improvement, Pupykina et al. [7] survey studies.
investigated memory management methods in cloud This research intends to close this gap by offering a
computing, and Xu et al. [6] assessed technologies for thorough analysis of cloud computing methods for DNN
managing virtual machine performance. Offloading deployment, along with considerations of difficulties and
technologies for job optimization across cloud and edge potential future research areas. The article's remaining
systems were covered by Wang et al. [8]. In order to address sections are structured to present different DNN
problems and look toward the future, Xu et al. [6] provide mechanisms, the need for cloud computing, popular cloud
an overview of computational distribution strategies for platforms for deploying DNNs, specific DNN applications
managing virtual machines in the cloud. Deep implemented in cloud systems, difficulties in current DNN
reinforcement learning-based cloud resource scheduling deployments using cloud computing systems, and
solutions were discussed by Zhou et al. [9] and Feng et al. opportunities for improving current DNN deployments on
[10]. cloud systems. Finally, a summary of the findings in brings
Regarding applications, Bera et al. [12] reviewed cloud the article to a close.
computing applications in smart grids, while Khan et al.

Fig.2: Deep learning architectures in emerging cloud computing architectures

(Source: Fatsuma Jauro, 2020. Deep learning architectures in emerging cloud computing architectures. Volume 96, 106582,
ISSN 1568-4946, https://doi.org/10.1016/j.asoc.2020.106582)

Computational Complexities in Deep Learning phases both need a considerable time commitment. This
A single deep neural network (DNN) consists of a section explores several popular DNN architectures, such as
significant number of parameters, demanding a large multilayer perceptron’s (MLPs), convolutional neural
amount of storage memory. A DNN's training and execution networks (CNNs), and graph neural networks (GNNs), all
This article can be downloaded from here: www.ijaems.com 10
©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

of which have intricate architectural designs and a wide units [30]. RNNs' hidden state, which is used for iterative
range of DNN parameters. Because DNN training requires processing, catches and holds onto data all the way through
extensive computation, it is impractical to use a single the sequence. RNNs have the ability to learn by storing,
isolated computer for this purpose, which highlights the retrieving, and using historical data to make predictions.
vital role that cloud computing plays in aiding DNN Model interactions at different temporal scales endow
training. memory. Information from all earlier steps is captured by
An extensively discussed neural network is the multilayer the aforementioned concealed state. In order to produce the
perceptron [24–26]. Each of MLP's levels, which include an output at a particular step in the sequence, the trained RNN
input layer, hidden layers, and an output layer, has a can combine the input sequence and the hidden state.
collection of perceptron components, also known as Despite being created more than a decade ago, RNNs still
neurons. An MLP with two hidden layers, an input layer, have issues with memory storage and computational time
and an output layer is shown in Figure 3. limits [31].
Figure 4 depicts a typical CNN with input/output layers,
hidden layers, and a fully connected network [41,42].
Convolution, activation, and pooling layers are all included
in a hidden layer. The activation function assists in learning
nonlinear input patterns from the convolution output, the
pooling layer consolidates the outputs of the activation
function into a single value, and the convolution layer
retrieves input features within this layer. To make
categorization easier, relevant features are retrieved after
numerous convolution and pooling procedures.

Fig.3: Multi-Layer Perceptron’s Typical Topology

Longer computational times are required to optimize an

MLP when number of layers are higher. Multiple apps have
been created as a result of using MLP in the cloud. For
instance, the study described in [28] concentrated on Fig.4: Typical Convolutional Neural Network (CNN)
developing a forecasting model that made use of multiple (Source:https://www.analyticsvidhya.com/blog/2022/01/co
input variables generated from several daily basic food price nvolutional-neural-network-an-overview/)
kinds. Using the Amazon Cloud Services infrastructure, this
model sought to forecast the Surabaya consumer price
index. The multilayer perceptron technique was used in the A CNN typically consists of millions of network
study to build a prediction system with a hidden layer, an parameters, each of which must be determined over the
course of a long period of time. In particular, ShuffleNet,
epoch, and a set number of neurons. Similar to this, a
GoogLeNet, DenseNet, ResNet, AlexNet, VGGNet, and
different study [29] divided regions afflicted by cancer and
ConvNet are a few CNN models that include millions of
stored relevant information in the cloud using transfer
learning-based cancer segmentation (TL-CAN-Seg) network parameters and need a lot of training time. For
technology. A unique MLP and an altered Levenberg- instance, it takes more than two weeks to train ConvNet on
a machine with four NVIDIA Titan Black GPUs.
Marquardt (LM) algorithm were used to interpret complex
Additionally, OpenAI's CNN, also known as neural
picture patterns and accomplish precise classification of
architecture search (NAS), requires six months to train
areas affected by breast cancer, improving the accuracy of
breast cancer diagnosis. when using 8 P100s in parallel scaling and has a remarkable
175 billion parameters for natural language processing.
Recurrent neural networks (RNNs) are more effective than
While purchasing many computers for a single CNN proves
MLPs at handling temporal data, including text and
economically inefficient, training a big CNN on a sole
sequentially correlated time series. The result from the
previous phase in the sequence serves as the input for the computer is prohibitively time-consuming and unfeasible.
Therefore, the use of CNNs in cloud infrastructure has been
next step in RNNs' unique versions of neural processing
proven in a number of applications [49-56].
This article can be downloaded from here: www.ijaems.com 11
©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

GNN is a result of graph representation learning [57–59], by offering substantial computing power and abundant data
which entails transforming and comprehending nodes and storage, eventually helping customers using DNNs in
edges in a graph into a continuous space with fewer intense applications [82].
dimensions. GNN treats non-Euclidean domains using The structure and composition of cloud data centers are
elaborate data structures that reflect the relationships outlined in the next section. The next part discusses
between these entities [60]. This is true even if it works with frequently used commercial cloud computing platforms for
Euclidean 1-D sequences like texts and 2-D grids like the deployment of DNN and provides an overview of public
images [61]. In GNNs, the underlying data structure is or volunteer cloud computing platforms. The section also
represented by a graph with the formula G = (V, E), where discusses frequently used cloud streaming systems, offering
V denotes the set of vertices or nodes and E denotes the light on how data streaming is implemented there.
edges joining them. It is possible for the relationship (u, v) Researchers studying deep learning (DL) who are looking
E to be symmetric or asymmetric. Graphs can be for reliable, affordable, and quick computing platforms for
homogenous, like social networks like Facebook friends, or DNN development will find this comprehensive material to
diverse, like knowledge graphs. The incorporated elements be of particular value.
or graphic topologies may also alter dynamically over time.
- Cloud data centers
GNNs' exponential node expansion increases
Data storage and computing are handled by cloud data
computational complexity and memory use significantly
centers or remote clouds, including backhaul and core
[62]. Notably, some sizable GNNs, like GemNet-XL,
networks [83]. A typical cloud computing architecture,
include billions of parameters [69]. Existing GNN models
shown in Figure 5, is made up of cloud users, internet
are judged insufficient for large-scale graphs that
network providers, and cloud service providers. Users
incorporate intricate topologies because they have only
provide computational data over network service providers,
been evaluated on small graphs [70]. Examples of GNN
which servers then receive. This data is processed using
implementation in cloud infrastructure have shown that
cloud resources, ensuring that users have enough access to
graph analysis is scalable and effective in a variety of
a shared pool of resources in response to their requests. This
applications, such as recommender systems, traffic flow
use of cloud resources makes it possible to offer adaptable
prediction, industrial IoT, privacy preservation, and matrix
processing power and storage, which eventually helps well-
completion [71–76].
known cloud-based companies like Amazon and Google
It is essential to comprehend the causes of latency in the Cloud stay profitable [84, 85].
training and inference phases of various DNN architectures
if you want to deploy them in the cloud with the least
amount of delay possible. Using methods like dilated
convolutions in CNNs, for instance, one can increase the
network's receptive field without adding more parameters
or layers, which lowers computational cost and inference
delay [77,78]. The use of backpropagation gradients
through time, which can be computationally expensive, is
not necessary with randomization-based learning
techniques like echo state networks (ESN) [79].
Additionally, GNN and CNN pruning strategies can reduce
the amount of parameters and computations needed, Fig.5: Typical Cloud Data Center Architecture
resulting in quicker inference times [80,81]. (Source: Elmirghani. 2018. GreenTouch GreenMeter Core
Cloud computing architectures for deep learning based Network Energy-Efficiency Improvement Measures and
applications Optimization. Journal of Optical Communications and
Networking. 10. 10.1364/JOCN.10.00A250)
Because deep neural network (DNN) architectures are
complex and demand a large number of parameters, training
and execution periods are prolonged. As a result, it is Users' requests for calculations are spread across a variety
impossible to train or deploy DNN using a single standalone of cloud platforms with numerous data centers [86].
computer. Cloud computing offers a practical answer to this Resource-intensive computations are made possible by
problem for these kinds of resource-intensive computations. resource sharing within and between data centers. To
Cloud computing meets the demanding computational increase processing power, a distributed cloud can also be
needs of several DNN implementations and training tasks connected to hybrid clouds, public clouds, and edge
This article can be downloaded from here: www.ijaems.com 12
©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

computing installations. User requests can be distributed deal with these limitations and improve cloud computing's
among close-by data centers in an effort to reduce data capabilities. For instance, Proteus, an elastic PS framework
transmission delay. The architecture of a distributed cloud created to scale up training on public clouds, was introduced
consists of numerous sub-clouds. Based on the availability by Harlap et al. [89]. The framework dynamically assigns
of resources, a central controller within the distributed cloud PSs and personnel using three transitional stages,
distributes computing workloads among various sub- maximizing cost reductions, particularly when temporary
clouds. revocable resources become available. Litz [90] created
- DL in the cloud logical executors, which map physical nodes to control the
executor states of specific applications, in an additional
The availability of customizable computing resources, a
effort to increase elasticity. This method also includes
critical component for a variety of DL workloads
micro-tasks for determining dependencies and assigning
characterized by different degrees of compute needs across
micro-tasks accordingly.
distinct activities and datasets, is one of the main advantages
of cloud computing. Numerous cloud providers offer Advanced learning frameworks
services to meet these needs, such as the auto-scaling Deep neural network (DNN) training in cloud environments
features of Amazon EC2 and the scale-up and scale-out will be more dependable and effective thanks to proposed
capabilities of Microsoft Azure. These services make it improvements in learning frameworks. For workload
possible to run DL workloads efficiently even while scheduling and dynamic resource scaling, the DL-driven
working with limited cloud resources. framework DL2 [91] combines supervised learning and
Parameter server reinforcement learning techniques. The DNN is initially
taught offline to assimilate resource allocation patterns from
To enable the scalability of distributed machine learning
prior judgments, and reinforcement learning is then used to
applications within cloud data centers, parameter servers
improve the training of the DNN. An exploitation-
(PSs) have been developed [87]. To train deep neural
exploration method was used by Chen et al. [92] to propose
networks, PSs have been incorporated into a number of deep
the dynamic PS load distribution scheme known as PSLD.
learning frameworks, including TensorFlow and MXNet.
The plan consists of three stages: gathering data on each PS,
Even with developments, total reliability cannot be assured
having workers create performance profiles, and choosing
because cloud data centers may have server outages. To
PSs based on performance indicators and communication
ensure the orderly completion of learning tasks in such
time.
contexts, preventative procedures are required to handle
job-sharing and backups [88]. The collective, which At the same time, Wang et al. [93] suggested the elastic
consists of a server group and a worker group, is the center parameter server (EPS), a simple solution that enables
of the PS framework. Both have job schedulers and worker dynamic resource allocation and deallocation for increased
nodes, which work together in DNN training, while the resource use and training efficiency. To improve scalability
latter also includes a manager and server nodes. Consistent and optimize resource utilization, this strategy incorporates
hashing is used to express shared parameters as vectors of heuristic scheduling modes like incoming job scheduling
(key, value). Operational data created by nodes is and running job scheduling. Additionally, more specialized
communicated to the server, which then distributes global frameworks have been created, with a particular emphasis
information to each node. This framework provides on DL workloads on private clouds. In order to reduce
flexibility in guaranteeing consistent data when the waiting times, Hu et al. [94 - 95] implemented an
algorithm is not sensitive to data inconsistencies, increasing optimization technique based on the idea of training
the reliability of the PS framework. It allows asynchronous progress and integer programming to handle resource
tasks and dependencies by initiating the necessary methods. scaling challenges particular to AWS and Huawei clouds.

Due to its widespread connectivity with a sizable and - Data streaming for the cloud
dynamic resource pool, the PS architecture is more suitable The continuous streaming of real-time data is essential for
with heterogeneous production data centers and public supporting deep learning (DL) applications in various
clouds than alternative techniques like AllReduce5. But the sensor networks and control systems, such as those found in
original PS architecture has some drawbacks, such as autonomous vehicles or smart grids. For activities like
imbalances, elasticity constraints, and static parameter recognition or decision-making, it is essential to acquire
assignments. It is not possible to incorporate more available measurements or data as soon as possible. Lack of access to
resources into ongoing training activities, and workload the most recent data versions may jeopardize system
distribution across nodes frequently does not maximize functionality and lower safety standards. Given the huge
resource capacity. Several strategies have been put out to amounts of data volumes contained in each sample, it is
This article can be downloaded from here: www.ijaems.com 13
©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

impractical to retrain deep neural networks (DNNs) using a automatically incorporates fuzzy rules from data streams,
standalone computer, highlighting the necessity of data using a simplification procedure to merge unnecessary
streaming in a cloud setting. hidden layers and control network growth.
Cloud data streaming platforms Results from experiments show that this approach
Platforms for streaming cloud data have found use in successfully limits network size while maintaining
government and academic settings for the analysis of data performance standards. In addition, Nguyen et al. [105]
gathered by sensor networks. To make geographic data have developed a sensor network for gathering maritime
processing easier, the Southern California Earthquake data, using a deep recurrent neural network combined with
Center, for instance, has set up a geophysical sensor streaming data to monitor fishing activities, spot smuggling,
network [96]. This network's tens of thousands of sensors forecast maritime pollution, and improve maritime traffic
continuously and quickly sample data. In order to better safety and security in real time. This comprehensive method
understand climate change and create systems for predicting effectively handles noisy and infrequently sampled data in
earthquakes and inland flooding, geospatial data must be maritime environments by combining latent variable
collected. Furthermore, to study seismic and hydrological modeling and data streaming to capture key elements within
features in North America, the Geodesy Advancing maritime dynamics.
Geosciences and EarthScope (GAGE) GPS network uses In order to provide real-time predictions, DNN designs in
information from more than a thousand GPS sensors [97]. cloud environments must incorporate minimal inference
Similar to this, the US National Science Foundation has latencies. Smaller DNN designs or the use of accelerators,
provided funding for the creation of a worldwide sensor together with the storage of data features in low-latency
network that will largely be used to study climate change storage locations for offline precomputing predictions, can
and the cycling of carbon [98]. improve serving latency. Additionally, essential for
Data-streaming approaches adjusting to newly streamed data and improving DNN
performance over time is the implementation of incremental
Given that many deep neural networks (DNNs) operate in
training [106]. Using model artifacts from well-known,
dynamic contexts, it is essential that they constantly absorb
publicly accessible DNNs, this entails routine updates of
new information or undergo retraining. A data-streaming
DNNs based on fresh streams of data, enabling updates
technique has been developed to handle this issue and
without the requirement for retraining from scratch.
determine whether streaming data is required for changing
DNN parameters [101]. This method incorporates a
strategic trade-off between training expenses and II. APPLICATIONS OF DNNS IN THE CLOUD
performance to decide if DNNs need to be updated. Its use Deep neural network (DNN) deployment in the cloud has
with TensorFlowOnSpark for three online learning become widely used in a wide range of applications. The
workloads has shown a decrease in total processing time. sections that follow provide an overview of a few of these
Similar to this, Ashfahani et al. [102] have suggested a data- applications, including wireless capsule endoscopy, travel,
streaming approach to modify network topologies in natural language processing (NLP), business intelligence
response to fresh input. This method makes it easier for (BI), cybersecurity, anomaly detection, and mobile-cloud-
network nodes to grow and shrink dynamically, improving assisted implementations. Tables that give a brief synopsis
performance while reducing complexity. In comparison to of the application's content follow these subsections. These
previous approaches, comparative evaluations on standard descriptions accurately identify the complex research issues
datasets have shown improved network performance and posed by the combination of cloud systems and DNNs.
decreased network complexity [102].
- Natural language processing
A similar advancement is the introduction of an incremental
The application of artificial intelligence (AI) to the interplay
high-order DL model by Li et al. [103] that is designed to
of computers and human language is known as natural
adapt to high-frequency data streams. The strategy
language processing (NLP). It entails the creation of
efficiently minimizes adaptation time by translating data
computational models and algorithms that provide
into a high-order tensor space, and then uses first-order
computers the ability to effectively comprehend, decode,
approximations to reduce the time-consuming parameter
and produce human language. NLP includes a wide range
incrementation frequently associated with iterative
of activities, including sentiment analysis, speech
procedures. DNNs may now adapt to dynamic situations
recognition, language translation, natural language
better than traditional iterative approaches, effectively
comprehension, and more. Its uses span from text analysis
satisfying real-time needs. A unique fuzzy neural network
has been introduced by Pratama et al. [104] that
This article can be downloaded from here: www.ijaems.com 14
©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

tools and chatbots to virtual assistants and language "smart" tourism industry, with the goal of providing
translation programs [107 - 115]. passengers with knowledgeable and personalized services.
- Business intelligence Applying sophisticated intelligent approaches for analysis
is necessary when working with datasets that are so diverse,
The technology, systems, and procedures utilized for the
detailed, and dispersed [157 – 176].
gathering, integrating, analyzing, and presenting of business
information are referred to as business intelligence (BI). - Remote medical diagnosis
Utilizing a range of software tools, it entails gathering The use of wireless capsule endoscopy (WCE) has
information from internal and external sources and turning significantly increased over the past two years. These
it into insights that can be used to inform strategic and methods provide a level of internal human visibility for
tactical business choices. Data mining, data analysis, diagnostic purposes that is comparable to standard
querying, and reporting are just a few of the tasks included endoscopy. These technologies were initially presented in
in business intelligence (BI), which frequently uses metrics 2000, and after going through clinical studies, they were
and key performance indicators (KPIs) to assess an given the go-ahead by the Food and Drug Administration in
organization's or business's success. By offering historical, 2001. Notably, these technologies provide improved
current, and predictive insights of corporate operations, BI's portability and have numerous uses in the delivery of
primary goal is to enable better business decision-making systemic biologics and healthcare services [177 – 190].
[117 – 120]. - Mobile-cloud-assisted applications
- Cybersecurity A developing technology with a wide range of possible
The practice of defending computer systems, networks, applications, the idea of mobile cloud-assisted applications.
programs, and data from online attacks, harm, or illegal The ability to transfer tasks to cloud servers, hence
access is known as cybersecurity. In order to protect prolonging the system's operational lifespan, is the main
information technology (IT) systems and networks from goal of mobile cloud computing (MCC). The computing
theft, damage, or disruption while maintaining the integrity, strain on mobile devices like smartphones, tablets, and
confidentiality, and availability of data, security policies iPads is also lessened by MCC. To ensure the effectiveness
and safeguards must be put in place. Malware, ransomware, of the job offloading inference engine and to ease resource
phishing scams, hacking, and other types of cybercrime are restrictions on smartphones, which often have much less
just a few of the risks that cybersecurity seeks to protect processing power compared to older approaches, an
against. It includes a wide range of tactics, tools, and best extensive evaluation based on traced data is carried out [195
practices for preventing, spotting, and dealing with digital – 197].
threats and security breaches [121 - 135]. Before using Deep Neural Networks (DNNs), it is urgently
- Anomaly detection necessary to remove unnecessary and redundant frames
Finding patterns or occurrences within a dataset that during the Wireless Capsule Endoscopy (WCE) operation
drastically depart from expected behavior is known as in order to prioritize the video content. However, there are
anomaly detection. Finding anomalies, outliers, or some significant difficulties with video prioritizing in
abnormalities that deviate from the predicted patterns or WCE, especially when there are limited resources and
trends in the data entails using statistical analysis, machine processing capabilities. Because of this, the integration of
learning methods, and data mining approaches. The MCC helps to offer affordable storage, robust
identification of anomalous or suspicious activity is critical computational power, and software services [198 – 199].
for assuring the integrity, security, and effectiveness of
operations in a number of different sectors, including III. CHALLENGES AND FUTURE
cybersecurity, fraud detection, system health monitoring, DIRECTIONS
and industrial quality control [140 – 154].
Cloud-based Deep Neural Networks (DNNs) have been
- Travel constructed for a variety of applications requiring extensive
AI efficiently organizes and processes large datasets big data analysis and high-performance computing, making
produced by both clients and service providers in the tourist use of the significant computational power and data storage
and hospitality industries. Notably, a sizeable portion of capacities provided by cloud platforms. Despite the fact that
pertinent data used in the tourism sector comes from GPS DNNs are effective tools for pattern recognition, they pose
applications and is frequently linked to social media, a number of research obstacles, including issues with
Internet of Things (IoT), and site traffic statistics. These energy consumption, the length of training and execution,
huge databases are managed within the framework of the data security, and cloud compatibility [203 – 275].
This article can be downloaded from here: www.ijaems.com 15
©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

- Energy efficiency [2] Q. Yan, F.R. Yu, Q.X. Gong, J.Q. Li, Software-defined
- Training cost networking (SDN) and distributed denial of service (DDoS)
- Scalability attacks in cloud computing environments: a survey, some
research issues, and challenges, IEEE Communications
- Data security
Surveys and Tutorials 18 (1) (2016) 602–622.
- Privacy awareness
[3] S.L. Nita and M.I. Mihailescu, On artificial neural network
- Cloud interoperability used in cloud computing security – a survey, in Proceedings
- Learning from non-stationary data: retraining of the International Conference of Electronics, Computers
efficiency and adaptation and Artificial Intelligence, 2018.
- Elastic implementations of deep learning models [4] P.J. Sun, Privacy protection and data security in cloud
and flexible resource allocation computing: a survey, challenges, and solutions, IEEE
- Deep reinforcement learning Access, vol. 7, pp. 147 420–147 452, 2019.
- 3D vision applications [5] K. Gai, J. Guo, L.H. Zhu, S. Yu, Blockchain meets cloud
computing: a survey, IEEE Transactions on Neural
- Optimization of DNNs
Networks and Learning Systems 22 (3) (2022) 2009–2030.
[6] F. Xu, F.M. Liu, H. Jin, and A.V. Vasilakos, Managing
IV. CONCLUSIONS performance overhead of virtual machines in cloud
computing: A survey, state of the art, and future directions,
DNNs have a wide range of applications in numerous real- Proceedings of the IEEE, vol. 102, no. 1, pp. 11–31, 2014.
world fields. Deploying DNNs on solitary stand-alone PCs [7] A. Pupykina and G. Agosta, Survey of memory management
or mobile devices, however, is frequently problematic for techniques for HPC and cloud computing, IEEE Access,
huge data storage and analysis applications due to their vol. 7, pp. 167 351–167 373, 2019.
computational complexity and the enormous number of [8] B. Wang, C.C. Wang, W.W. Huang, Y. Song, X.Y. Qin, A
parameters necessary in training. As a result, the use of survey and taxonomy on task offloading for edge-cloud
computing, IEEE Access 8 (2020) 186 080–186 101.
DNNs in cloud computing systems has attracted a lot of
[9] G. Zhou, W. Tian, R. Buyya, Deep reinforcement learning-
interest. In the beginning, this review article lays out the
based methods for resource scheduling in cloud computing:
justification for using and training DNNs in cloud-based A review and future directions, Journal of Cloud Computing
systems. It then digs into the intricate computations of 11 (2022), paper number 3.
popular DNNs, such as MLP, CNN, RNN, and GNN, [10] Y. Feng and F. Liu, Resource management in cloud
highlighting their high parameter and FLOPs needs. computing using deep reinforcement learning: A survey, in
Proceedings of the 10th Chinese Society of Aeronautics and
The research also provides a thorough analysis of volunteer
Astronautics Youth Forum, 2023, pp. 635–643.
and public cloud computing platforms that have effectively
[11] A.R. Khan, M. Othman, S.A. Madani, S.U. Khan, A survey
included and applied DNNs. Researchers and software of mobile cloud computing application models, IEEE
developers can use this information to choose the best cloud Communications Surveys and Tutorials 16 (1) (2014) 393–
computing platform for their DNN-focused apps. The paper 413.
also provides information on a number of application fields, [12] S. Bera, S. Misra, J.J.P.C. Rodrigues, Cloud computing
including NLP, BI, cybersecurity, anomaly detection, and applications for smart grid: a survey, IEEE Transactions on
travel, which have recently reaped significant advantages Parallel and Distributed Systems 26 (5) (2015) 1477–1494.
from the integration of DNN in cloud computing. It outlines [13] K. Cao, S.Y. Hu, Y. Shi, A.W. Colombo, S. Karnouskos, X.
Li, A survey on edge and edge-cloud computing assisted
the key difficulties involved in this approach while
cyber-physical systems, IEEE Transactions on Industrial
highlighting the benefits and efficiency of installing DNNs
Informatics 17 (11) (2021) 7806–7819.
in cloud-based applications. [14] D. Soni, N. Kumar, Machine learning techniques in emerging
The research also suggests potential possibilities to improve cloud computing integrated paradigms: A survey and
current deployments using cloud computing platforms and taxonomy, Journal of Network and Computer Applications
DNNs. This thorough overview study is projected to be a 205 (2022).
[15] T. Khana, W.H. Tiana, R. Buyya, Machine learning (ml)-
useful resource for researchers and programmers interested
centric resource management in cloud computing: A review
in successfully implementing DNNs on cloud computing
and future directions, Journal of Network and Computer
platforms. Applications 204 (2022).
[16] A. Saiyeda, M.A. Mir, Cloud computing for deep learning
analytics: a survey of current trends and challenges,
REFERENCES
International Journal of Advanced Research in Computer
[1] E. Strubell, A. Ganesh, and A. McCallum, Energy and policy Science 8 (2) (2017).
considerations for deep learning in nlp, in: Annual Meeting [17] P.S. Priya, P. Malik, A. Mehbodniya, V. Chaudhary, A.
of the Association for Computational Linguistics, 2019. Sharma, and S. Ray, The relationship between cloud

This article can be downloaded from here: www.ijaems.com 16

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

computing and deep learning towards organizational [31] Z.C. Lipton, J. Berkowitz, and C. Elkan, A critical review of
commitment, in: Proceedings of the 2nd International recurrent neural
Conference on Innovative Practices in Technology and networks for sequence learning, arXiv preprint arXiv:1506.00019,
Management, 2022. 2015.
[18] F. Benedetto and A. Tedeschi, Big data sentiment analysis for [32] R. Pascanu, T. Mikolov, and Y. Bengio, On the difficulty of
brand monitoring in social media streams by cloud training recurrent neural networks, in: Proceedings of
computing, in: Sentiment Analysis and Ontology International Conference on Machine Learning, 2013, pp.
Engineering. Springer, 2016, pp. 341–377. 1310–1318.
[19] S. Mohan, S. Mullapudi, S. Sammeta, P. Vijayvergia, and [33] P. Yazdanian, S. Sharifian, E2lg: a multiscale ensemble of
D.C. Anastasiu, Stock price prediction using news sentiment lstm/gan deep learning architecture for multistep-ahead
analysis, in: 2019 IEEE Fifth International Conference on cloud workload prediction, The Journal of Supercomputing
Big Data Computing Service and Applications 77 (2021) 11 052–11 082.
(BigDataService). IEEE, 2019, pp. 205–208. [34] Y.S. Patel, J. Bedi, MAG-D: A multivariate attention network
[20] S. Prasomphan, Improvement of chatbot in trading system for based approach for cloud workload forecasting, Future
smes by using deep neural network, in: 2019 IEEE 4th Generation Computer Systems (2023).
International Conference on Cloud Computing and Big Data [35] S. Ouhame, Y. Hadi, A. Ullah, An efficient forecasting
Analysis (ICCCBDA). IEEE, 2019, pp. 517–522. approach for resource utilization in cloud data center using
[21] D. Scheinert, A. Acker, L. Thamsen, M.K. Geldenhuys, and cnn-lstm model, Neural Computing and Applications 33
O. Kao, Learning dependencies in distributed cloud (2021) 10 043–10 055.
applications to identify and localize anomalies, in: [36] N. Tran, T. Nguyen, B.M. Nguyen, G. Nguyen, A multivariate
Proceedings of IEEE/ACM International Workshop on fuzzy time series resource forecast model for clouds using
Cloud Intelligence, 2021, pp. 7–12. lstm and data correlation analysis, Procedia Computer
[22] M.A. Elsayed, M. Zulkernine, Predictdeep: Security analytics Science 126 (2018) 636–645.
as a service for anomaly detection and prediction, IEEE [37] H.L. Leka, Z. Fengli, A.T. Kenea, A.T. Tegene, P. Atandoh,
Access 8 (2020) 45 184–45 197. and N.W. Hundera, A hybrid cnn-lstm model for virtual
[23] F. Jauro, H. Chiroma, A.Y. Gital, M. Almutairi, S.M. machine workload forecasting in cloud data center, in: 2021
Abdulhamid, J.H. Abawajy, Deep learning architectures in 18th International Computer Conference on Wavelet Active
emerging cloud computing architectures: Recent Media Technology and Information Processing
development, challenges and next research trend, Applied (ICCWAMTIP). IEEE, 2021, pp. 474–478.
Soft Computing Journal 96 (2020), paper number 106582. [38] Z. Ding, J. Wang, Y. Cheng, and C. He, Alice: A lstm neural
[24] J.-F. Chen, Q.H. Do, H.-N. Hsieh, Training artificial neural network based short-term power load forecasting approach
networks by a hybrid pso-cs algorithm, Algorithms 8 (2) in distributed cloud-edge environment, in: Journal of
(2015) 292–308. Physics: Conference Series, vol. 1624, no. 5. IOP
[25] A.A. Heidari, H. Faris, S. Mirjalili, I. Aljarah, M. Mafarja, Publishing, 2020, p. 052017.
Ant lion optimizer: theory, literature review, and application [39] T. Hussain, K. Muhammad, W. Ding, J. Lloret, S.W. Baik,
in multi-layer perceptron neural networks, Nature-Inspired V.H.C. de Albuquerque, A comprehensive survey of multi-
Optimizers (2020) 23–46. view video summarization, Pattern Recognition 109 (2021).
[26] Y.-S. Park, S. Lek, Artificial neural networks: multilayer [40] K. Muhammad, T. Hussain, J. Del Ser, V. Palade, V.H.C. De
perceptron for ecological modeling, Developments in Albuquerque, Deepres: A deep learning-based video
environmental modelling, vol. 28, ch. 7, Elsevier, 2016, pp. summarization strategy for resource constrained industrial
123–140. surveillance scenarios, IEEE Transactions on Industrial
[27] R. Qaddoura, M. Al-Zoubi, H. Faris, I. Almomani, et al., A Informatics 16 (9) (2019) 5938–5947.
multi-layer classification approach for intrusion detection in [41] K. Fukushima, S. Miyake, T. Ito, Neocognitron: A neural
iot networks based on deep learning, Sensors 21 (9) (2021) network model for a mechanism of visual pattern
2987. recognition, IEEE Transactions on Systems, Man and
[28] S. Zahara, S. Sugianto, Prediksi indeks harga konsumen Cybernetics 13 (5) (1983) 826–834.
komoditas makanan berbasis cloud computing [42] A. Krizhevsky, I. Sutskever, and G. Hinton, Imagenet
menggunakan multilayer perceptron, JOINTECS (Journal classification with deep convolutional neural network, in:
of Information Technology and Computer Science) 6 (1) Proceedings of the Conference of Neural Information
(2021) 21–28. Processing Systems, 2012, pp. 1106–1114.
[29] G.S.B. Jahangeer, T.D. Rajkumar, Cloud storage based [43] X. Zhang, X. Zhou, M. Lin, and J. Sun, Shufflenet: An
diagnosis of breast cancer using novel transfer learning with extremely efficient convolutional neural network for mobile
multi-layer perceptron, International Journal of System devices, in: Proceedings of The IEEE Computer Vision and
Assurance Engineering and Management (2022) 1–13. Pattern Recognition Conference, 2018.
[30] D. Mandic, J. Chambers, Recurrent neural networks for [44] C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. Reed, D.
prediction: learning algorithms, architectures and stability, Anguelov, D. Erhan, V. Vanhoucke, and A. Rabinovich,
Wiley, 2001. Going deeper with convolutions, in: Proceedings of The

This article can be downloaded from here: www.ijaems.com 17

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

IEEE Computer Vision and Pattern Recognition [58] P. Cui, X. Wang, J. Pei, W. Zhu, A survey on network
Conference, 2015. embedding, IEEE Transactions on Knowledge and Data
[45] G. Huang, Z. Liu, L. van der Maaten, and K.Q. Weinberger, Engineering 31 (5) (2018) 833–852.
Densely connected convolutional networks, in: Proceedings [59] W.L. Hamilton, R. Ying, and J. Leskovec, Representation
of The IEEE Computer Vision and Pattern Recognition learning on graphs: Methods and applications, in:
Conference, 2018. Proceedings of IEEE Data Engineering Bulletin, 2017.
[46] K. He, X. Zhang, S. Ren, J. Sun, Deep residual learning for [60] M.M. Bronstein, J. Bruna, Y. LeCun, A. Szlam, P.
image recognition, Microsoft Research, Technical Report Vandergheynst, Geometric deep learning: going beyond
(2015) [Online]. euclidean data, IEEE Signal Processing Magazine 34 (4)
Available:https://arxiv.org/pdf/1512.03385.pdf. (2017) 18–42.
[47] A. Krizhevsky, I. Sutskever, and G.E. Hinton, Imagenet [61] J. Zhou, G. Cui, S. Hu, Z. Zhang, C. Yang, Z. Liu, L. Wang,
classification with deep convolutional neural networks, in: C. Li, M. Sun, Graph neural networks: A review of methods
Proceedings of Conference on Neural Information and applications, AI Open 1 (2020) 57–81.
Processing Systems. IEEE, 2012. [62] F. Frasca, E. Rossi, D. Eynard, B. Chamberlain, M. Bronstein,
[48] K. Simonyan and A. Zisserman, Very deep convolutional and F. Monti, Sign: Scalable inception graph neural
networks for largescale image recognition, in: Proceedings networks, in: International Conference on Machine
of the International Conference on Learning Learning, 2020.
Representations, 2015. [63] W. Hu, M. Shuaibi, A. Das, S. Goyal, A. Sriram, J. Leskovec,
[49] A.D. Torres, H. Yan, A.H. Aboutalebi, A. Das, L. Duan, and D. Parikh, and C.L. Zitnick, Forcenet: A graph neural
P. Rad, Patient facial emotion recognition and sentiment network for large-scale quantum calculations, in:
analysis using secure cloud with hardware acceleration, in: Proceedings of The Conference on Neural Information
Computational Intelligence for Multimedia Big Data on the Processing Systems, 2021.
Cloud with Engineering Applications. Elsevier, 2018, pp. [64] J. Klicpera, S. Giri, J.T. Margraf, and S. Günnemann, Fast and
61–89. uncertainty aware directional message passing for non-
[50] A. Makkar, U. Ghosh, P.K. Sharma, Artificial intelligence equilibrium molecules, in: Proceedings of The Conference
and edge computing enabled web spam detection for next on Neural Information Processing Systems, 2020.
generation iot applications, IEEE Sensors Journal 21 (22) [65] M. Shuaibi, A. Kolluru, A. Das, A. Grover, A. Sriram, Z.
(2021) 25 352–25 361. Ulissi, and C.L. Zitnick, Rotation invariant graph neural
[51] D. Selvapandian, R. Santhosh, Deep learning approach for networks using spin convolutions, in: Proceedings of The
intrusion detection in iot-multi cloud environment, Conference on Neural Information Processing Systems,
Automated Software Engineering 28 (2) (2021) 1–17. 2021.
[52] D. Scheinert and A. Acker, Telesto: A graph neural network [66] J. Klicpera, F. Becker, and S. Günnemann, Gemnet: Universal
model for anomaly classification in cloud services, in: directional graph neural networks for molecules, in:
Proceedings of International Conference on Service- Proceedings of The Conference on Neural Information
Oriented Computing. Springer, 2020, pp. 214–227. Processing Systems, 2021.
[53] W. Ullah, A. Ullah, I.U. Haq, K. Muhammad, M. Sajjad, S.W. [67] Available:
Baik, CNN features with bi-directional LSTM for real-time https://blog.twitter.com/engineering/en_us/topics/insights/2
anomaly detection in surveillance networks, Multimedia 020/graph-ml-at-twitter.
Tools and Applications 80 (11) (2021) 16 979–16 995. [68] A. Sriram, A. Das, B.M. Wood, S. Goyal, and C.L. Zitnick,
[54] H. Saito, T. Aoki, K. Aoyama, Y. Kato, A. Tsuboi, A. Towards training billion parameter graph neural networks
Yamada, M. Fujishiro, S. Oka, S. Ishihara, T. Matsuda, et for atomic simulations, in: Proceedings of The International
al., Automatic detection and classification of protruding Conference on Learning Representations, 2022.
lesions in wireless capsule endoscopy images based on a [69] J. Gasteiger, F. Becker, and S. Gunnemann, Gemnet:
deep convolutional neural network, Gastrointestinal Universal directional graph neural networks for molecules,
endoscopy 92 (1) (2020) 144–151. in: Proceedings of The Conference on Neural Information
[55] G. Shu, W. Liu, X. Zheng, J. Li, If-cnn: Image-aware Processing Systems, 2022.
inference framework for cnn with the collaboration of [70] Z. Wu, S. Pan, F. Chen, G. Long, C. Zhang, S.Y. Philip, A
mobile devices and cloud, IEEE Access 6 (2018) 68 621–68 comprehensive survey on graph neural networks, IEEE
633. Transactions on Neural Networks and Learning Systems 32
[56] C.-H. Chen, C.-R. Lee, W.C.-H. Lu, Smart in-car camera (1) (2020) 4–24.
system using mobile cloud computing framework for deep [71] R. Ying, R. He, K. Chen, P. Eksombatchai, W.L. Hamilton,
learning, Vehicular Communications 10 (2017) 84–90. and J. Leskovec, Graph convolutional neural networks for
[57] B. Abu-Salih, M. Al-Tawil, I. Aljarah, H. Faris, P. web-scale recommender systems, in: Proceedings of the
Wongthongtham, K.Y. Chan, A. Beheshti, Relational 24th ACM SIGKDD international conference on knowledge
learning analysis of social politics using knowledge graph discovery & data mining, 2018, pp. 974–983.
embedding, Data Mining and Knowledge Discovery (2021) [72] L. Ge, Y. Jia, Q. Li, X. Ye, Dynamic multi-graph convolution
1–40. recurrent neural network for traffic speed prediction, Journal
of Intelligent & Fuzzy Systems no. Preprint (2023) 1–14.

This article can be downloaded from here: www.ijaems.com 18

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

[73] R. Yumlembam, B. Issac, S.M. Jacob, and L. Yang, Iot-based distributed machine learning with the parameter server, in:
android malware detection using graph neural network with Proceedings of Symposium on Operating Systems Design
adversarial defense, IEEE Internet of Things Journal, 2022. and Implementation, 2014, pp. 583–598.
[74] Y. Li, S. Xie, Z. Wan, H. Lv, H. Song, Z. Lv, Graph-powered [89] A. Harlap, A. Tumanov, A. Chung, G.R. Ganger, and P.B.
learning methods in the internet of things: A survey, Gibbons, Proteus: agile ML elasticity through tiered
Machine Learning with Applications 11 (2023). reliability in dynamic resource markets, in: Proceedings of
[75] S. Wang, Y. Zheng, X. Jia, Secgnn: Privacy-preserving graph the Twelfth European Conference on Computer Systems,
neural network training and inference as a cloud service, 2017, pp. 589–604.
IEEE Transactions on Services Computing (2023). [90] A. Qiao, A. Aghayev, W. Yu, H. Chen, Q. Ho, G.A. Gibson,
[76] R. v. d. Berg, T.N. Kipf, and M. Welling, Graph convolutional and E.P. Xing, Litz: Elastic framework for high-
matrix completion, arXiv preprint arXiv:1706.02263, 2017. performance distributed machine learning, in: Proceedings
[77] S. Pratiher, S. Chattoraj, D. Nawn, M. Pal, R.R. Paul, H. of Annual Technical Conference, 2018, pp. 631–644.
Konik, and J. Chatterjee, A multi-scale context aggregation [91] Y. Peng, Y. Bao, Y. Chen, C. Wu, C. Meng, W. Lin, Dl2: A
enriched mlp-mixer model for oral cancer screening from deep learning-driven scheduler for deep learning clusters,
oral sub-epithelial connective tissues, in: 2022 30th IEEE Transactions on Parallel and Distributed Systems 32
European Signal Processing Conference (EUSIPCO). IEEE, (8) (2021) 1947–1960.
2022, pp. 1323–1327. [92] Y. Chen, Y. Peng, Y. Bao, C. Wu, Y. Zhu, and C. Guo, Elastic
[78] Y. Li, K. Li, X. Liu, Y. Wang, L. Zhang, Lithium-ion battery parameter server load distribution in deep learning clusters,
capacity estimation–a pruned convolutional neural network in: Proceedings of the 11th ACM Symposium on Cloud
approach assisted with transfer learning, Applied Energy Computing, 2020, pp. 507–521.
285 (2021). [93] S. Wang, A. Pi, X. Zhou, Elastic parameter server:
[79] P.N. Suganthan, R. Katuwal, On the origins of randomization- Accelerating ML training with scalable resource scheduling,
based feedforward neural networks, Applied Soft IEEE Transactions on Parallel and Distributed Systems 33
Computing 105 (2021). (5) (2021) 1128–1143.
[80] T. Le-Cong, H.J. Kang, T.G. Nguyen, S.A. Haryono, D. Lo, [94] L. Hu, J. Zhu, Z. Zhou, R. Cheng, X. Bai, and Y. Zhang, An
X.-B.D. Le, and Q.T. Huynh, Autopruner: transformer- optimal resource allocator of elastic training for deep
based call graph pruning, in: Proceedings of the 30th ACM learning jobs on cloud, arXiv preprint arXiv:2109.03389,
Joint European Software Engineering Conference and 2021.
Symposium on the Foundations of Software Engineering, [95] T. Menouer, KCSS: kubernetes container scheduling strategy,
2022, pp. 520–532. The Journal of Supercomputing 77 (5) (2021) 4267–4293.
[81] J. Poyatos, D. Molina, A.D. Martinez, J. Del Ser, F. Herrera, [96] N. Nakata, J.P. Chang, J.F. Lawrence, P. Boue, Body wave
Evoprunedeeptl: An evolutionary pruning model for transfer extraction and tomography at long beach california with
learning based deep neural networks, Neural Networks 158 ambient noise interferometry, Journal of Geophysical
(2023) 59–82. Research: Solid Earth 120 (2) (2015) 1159–1173.
[82] D.T. Hoang, C. Lee, D. Niyato, P. Wang, A survey of mobile [97] D.A. Philips, C.P. Santillan, L.M. Wang, R.W. King, W.M.
cloud computing: architecture, applications, and Szeliga, T. Melbourne, M. Floyd, T.A. Herring, Plate
approaches, Wireless Communications and Mobile boundary observatory and related networks: GPS data
Computing 13 (2013) 1587–1611. analysis methods and geodetic products, Reviews of
[83] Q. Zhang, L. Cheng, R. Boutaba, Cloud computing: State-of- Geophysics 54 (2016) 759–808.
the-art and research challenges, Journal of Internet Services [98] T. Cowles, J. Delaney, J. Orcutt, R. Weller, The ocean
and Applications 1 (1) (2010) 7–18. observatories initiative: sustained ocean observing across a
[84] T. Dillon, C. Wu, and E. Chang, Cloud computing: Issues and range of spatial scales, Marine Technology Society Journal
challenges, in: Proceedings of the IEEE International 44 (6) (2010) 54–64.
Conference on Advanced Information Networking and [99] I. Foster, D.B. Gannon, Cloud Computing For Science And
Applications, 2010. Engineering, The MIT Press, 2017.
[85] J. Novet. (2021) How amazon’s cloud business generates [100] T. Akidau, R. Bradshaw, C. Chambers, S. Chernyak, R.J.
billions in profit. [Online]. Available: Fernandez-Moctezuma, R. Lax, S. McVeety, D. Mills, F.
https://www.cnbc.com/2021/09/05/how-amazon- Perry, E. Schmidt, and S. Whittle, The dataflow model: a
webservices-makes-money-estimated-margins-by- practical approach to balancing correctness, latency, and
service.html. cost in massive scale, unbounded, out-of-order data
[86] What is distributed cloud? 2022. processing, Proceedings of the VLDB ENdowment, vol. 8,
https://www.windriver.com/solutions/learning/distributed- no. 12, pp. 1792–1803, 2015.
cloud. [101] W. Rang, D.L. Yang, D.Z. Cheng, Y. Wang, Data life aware
[87] A. Smola and S. Narayanamurthy, An architecture for parallel model updating strategy for stream-based online deep
topic models, Proceedings of the VLDB Endowment, vol. 3, learning, IEEE Transactions on Parallel and Distributed
no. 1–2, pp. 703–710, 2010. Systems 32 (10) (2021) 2571–2581.
[88] M. Li, D.G. Andersen, J.W. Park, A.J. Smola, A. Ahmed, V. [102] A. Ashfahani, Autonomous deep learning: Incremental
Josifovski, J. Long, E.J. Shekita, and B.-Y. Su, Scaling learning of deep neural networks for evolving data streams,

This article can be downloaded from here: www.ijaems.com 19

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

in: Proceedings of the International Conference on Data [117] M.A. Khan, S. Saqib, T. Alyas, A.U. Rehman, Y. Saeed, A.
Mining Workshops, 2019, pp. 83–90. Zeb, M. Zareei, E.M. Mohamed, Effective demand
[103] Y.L. Li, M. Zhang, W. Wang, Online real-time analysis of forecasting model using business intelligence empowered
data streams based on an incremental high-order deep with machine learning, IEEE Access 8 (2020) 116 013–116
learning model, IEEE Access 6 (2018) 77 615–77 623. 023.
[104] M. Pratama, W. Pedrycz, G.I. Webb, Online real-time [118] B.M. Balachandran, S. Prasad, Challenges and benefits of
analysis of data streams based on an incremental high-order deploying big data analytics in the cloud for business
deep learning model, IEEE Transactions on Fuzzy Systems intelligence, Procedia Computer Science 112 (2017) 1112–
28 (7) (2020) 1315–1328. 1122.
[105] D. Nguyen, R. Vadaine, G. Hajduch, R. Garello, and R. [119] C. Moreno, R.A.C. González, E.H. Viedma, Data and
Fablet, A multi-task deep learning architecture for maritime artificial intelligence strategy: A conceptual enterprise big
surveillance using ais data streams, in: Proceedings of IEEE data cloud architecture to enable market-oriented
5th International Conference on Data Science and Advanced organisations, International Journal of Interactive
Analytics, 2018, pp. 331–340. Multimedia and Artificial Intelligence 5 (6) (2019) 7–14.
[106] S.S. Zhang, J.W. Liu, and X. Zuo, Adaptive online [120] C. Juarez and H. Afli, Online news analysis on cloud
incremental learning for evolving data streams, 2022. computing platform for market prediction. in Collaborative
[Online]. Available: https://arxiv.org/abs/1805.04754. European Research Conference, 2020, pp. 125–140.
[107] D.W. Otter, J.R. Medina, J.K. Kalita, A survey of the usages [121] P. Dixit and S. Silakari, ‘‘Deep learning algorithms for
of deep learning for natural language processing, IEEE cybersecurity applications: A technological and status
Transactions on Neural Networks and Learning Systems 32 review,” Computer Science Review, vol. 39, paper number
(2) (2021) 604–624. 100317, 2021
[108] W. Medhat, A. Hassan, H. Korashy, Sentiment analysis [122] M.M. Alani, Big data in cybersecurity: a survey of
algorithms and applications: A survey, Ain Shams applications and future trends, Journal of Reliable
Engineering Journal 5 (4) (2014) 1093–1113. Intelligent Environments (2021) 1–30.
[109] R. Feldman, Techniques and applications for sentiment [123] P. Podder, S. Bharati, M. Mondal, P.K. Paul, and U. Kose,
analysis, Communications of the ACM 56 (4) (2013) 82–89. Artificial neural network for cybersecurity: a
[110] R. Obiedat, L. Al-Qaisi, R. Qaddoura, O. Harfoushi, A. Al- comprehensive review, arXiv preprint arXiv:2107.01185,
Zoubi, An intelligent hybrid sentiment analyzer for personal 2021.
protective medical equipments based on word embedding [124] L. Gupta, T. Salman, A. Ghubaish, D. Unal, A.K. Al-Ali, R.
technique: The COVID-19 era, Symmetry 13 (12) (2021) Jain, Cybersecurity of multi-cloud healthcare systems: A
2287. hierarchical deep learning approach, Applied Soft
[111] D.A. Al-Qudah, A.-Z. Ala’M, P.A. Castillo-Valdivieso, H. Computing 118 (2022), paper number 108439.
Faris, Sentiment analysis for e-payment service providers [125] Y. Chai, Y. Zhou, W. Li, Y. Jiang, An explainable multi-
using evolutionary extreme gradient boosting, IEEE Access modal hierarchical attention model for developing phishing
8 (2020) 189 930–189 944. threat intelligence, IEEE Transactions on Dependable and
[112] R. Obiedat, O. Harfoushi, R. Qaddoura, L. Al-Qaisi, A. Al- Secure Computing 19 (2) (2022) 790–803.
Zoubi, An evolutionary-based sentiment analysis approach [126] B. Abu-Salih, D.A. Qudah, M. Al-Hassan, S.M. Ghafari, T.
for enhancing government decisions during COVID-19 Issa, I. Aljarah, A. Beheshti, and S. Alqahtan, An intelligent
pandemic: The case of jordan, Applied Sciences 11 (19) system for multi-topic social spam detection in
(2021) 9080. microblogging, arXiv preprint arXiv:2201.05203, 2022.
[113] R.O. Sinnott and S. Cui, Benchmarking sentiment analysis [127] F.J. Abdullayeva, Advanced persistent threat attack
approaches on the cloud, in: Proceedings of 2016 IEEE 22nd detection method in cloud computing based on autoencoder
International Conference on Parallel and Distributed and softmax regression algorithm, Array 10 (2021), paper
Systems, 2016, pp. 695–704. number 100067.
[114] M. Ghorbani, M. Bahaghighat, Q. Xin, F. Özen, [128] A. Vadariya and N.K. Jadav, A survey on phishing URL
Convlstmconv network: a deep learning approach for detection using artificial intelligence, in: Proceedings of
sentiment analysis in cloud computing, Journal of Cloud International Conference on Recent Trends in Machine
Computing 9 (1) (2020) 1–12. Learning, IoT, Smart Cities and Applications, 2021, pp. 9–
[115] M.R. Raza, W. Hussain, and J.M. Merigó, Long short-term 20.
memory-based sentiment classification of cloud dataset, in: [129] L. Tang, Q.H. Mahmoud, A survey of machine learning-
Proceedings of IEEE Innovations in Intelligent Systems and based solutions for phishing website detection, Machine
Applications Conference, 2021, pp. 1–6. Learning and Knowledge Extraction 3 (3) (2021) 672–694.
[116] G. Preethi, P.V. Krishna, M.S. Obaidat, V. Saritha, and S. [130] G. Li, P. Sharma, L. Pan, S. Rajasegarar, C. Karmakar, N.
Yenduri, Application of deep learning to sentiment analysis Patterson, Deep learning algorithms for cyber security
for recommender system on cloud, in: Proceedings of applications: A survey, Journal of Computer Security 29 (5)
International conference on computer, information and (2021) 447–471.
telecommunication systems. IEEE, 2017, pp. 93–97. [131] K. Sethi, R. Kumar, N. Prajapati, and P. Bera, Deep
reinforcement learning based intrusion detection system for

This article can be downloaded from here: www.ijaems.com 20

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

cloud infrastructure, in: Proceedings of the IEEE International Conference on Big Data. IEEE, 2020, pp.
International Conference on Communication Systems and 1301–1308.
Networks, 2020, pp. 1–6. [145] M. Lee, S. Hosseinalipour, C.G. Brinton, G. Yu, and H. Dai,
[132] M.N. Hossain, J. Wang, O. Weisse, R. Sekar, D. Genkin, B. A fast graph neural network-based method for winner
He, S.D. Stoller, G. Fang, F. Piessens, and E. Downing, determination in multi-unit combinatorial auctions, IEEE
Dependence-preserving data compaction for scalable Transactions on Cloud Computing, 2020.
forensic analysis, in: Proceedings of the 27th Security [146] M. Gao, Y. Li, and J. Yu, Workload prediction of cloud
Symposium, 2018, pp. 1723–1740. workflow based on graph neural network, in: Proceedings of
[133] M. Du, F. Li, G. Zheng, and V. Srikumar, Deeplog: Anomaly International Conference on Web Information Systems and
detection and diagnosis from system logs through deep Applications, 2021, pp. 169–189.
learning, in: Proceedings of the 2017 ACM SIGSAC [147] A. Rafiq, T.A. Khan, M. Afaq, and W.-C. Song, Service
Conference on Computer and Communications Security, function chaining and traffic steering in sdn using graph
2017, pp. 1285–1298. neural network, in: Proceedings of IEEE International
[134] T. Thilagam, R. Aruna, Intrusion detection for network Conference on Information and Communication
based cloud computing by custom rc-nn and optimization, Technology Convergence, 2020, pp. 500–505.
ICT Express 7 (4) (2021) 512–520. [148] Z. Yu, W. Liu, X. Liu, and G. Wang, Drag-jdec: A deep
[135] Y. Shen, E. Mariconti, P.A. Vervier, and G. Stringhini, reinforcement learning and graph neural network-based job
Tiresias: Predicting security events through deep learning, dispatching model in edge computing, in: Proceedings of
in: Proceedings of the 2018 ACM SIGSAC Conference on IEEE/ACM 29th International Symposium on Quality of
Computer and Communications Security, 2018, pp. Service. IEEE, 2021, pp. 1–10.
592–605. [149] T. Liang, X. Sheng, L. Zhou, Y. Li, H. Gao, Y. Yin, L. Chen,
[136] M. Ramilli, Malware training sets: a machine learning Mobile app recommendation via heterogeneous graph
dataset for everyone, Marco Ramilli Web Corner, 2016. neural network in edge computing, Applied Soft Computing
[137] N. Koroniotis, N. Moustafa, E. Sitnikova, B. Turnbull, vol. 103, paper number 107162 (2021).
Towards the development of realistic botnet dataset in the [150] A. Said, S.-U. Hassan, S. Tuarob, R. Nawaz, M. Shabbir,
internet of things for network forensic analytics: Bot-iot DGSD: Distributed graph representation via graph statistical
dataset, Future Generation Computer Systems 100 (2019) properties, Future Generation Computer Systems 119
779–796. (2021) 166–175.
[138] Z. Yang, D. Yang, C. Dyer, X. He, A. Smola, and E. Hovy, [151] S. Chen, J. Dong, P. Ha, Y. Li, S. Labi, Graph neural
Hierarchical attention networks for document classification, network and reinforcement learning for multi-agent
in: Proceedings of the 2016 conference of the North cooperative control of connected autonomous vehicles,
American chapter of the association for computational Computer-Aided Civil and Infrastructure Engineering 36 (7)
linguistics: human language technologies, 2016, pp. 1480– (2021) 838–857.
1489. [152] Y. Gao, X. Gu, H. Zhang, H. Lin, and M. Yang, ‘‘Runtime
[139] N. Moustafa and J. Slay, Unsw-nb15: a comprehensive data performance prediction for deep learning models with graph
set for network intrusion detection systems (unsw-nb15 neural network,” Technical Report MSR-TR-2021-3.
network data set), in: 2015 military communications and Microsoft, Tech. Rep., 2021.
information systems conference (MilCIS). IEEE, 2015, pp. [153] W. Ullah, A. Ullah, T. Hussain, Z.A. Khan, S.W. Baik, An
1–6. efficient anomaly recognition framework using an attention
[140] L. Bergman and Y. Hoshen, Classification based anomaly residual lstm in surveillance videos, Sensors 21 (8) (2021)
detection for general data, in: Proceedings of International 2811.
Conference on Learning Representations, 2020. [154] W. Ullah, A. Ullah, T. Hussain, K. Muhammad, A.A.
[141] Y. Wu, H.-N. Dai, and H. Tang, Graph neural networks for Heidari, J. Del Ser, S.W. Baik, V.H.C. De Albuquerque,
anomaly detection in industrial internet of things, IEEE Artificial intelligence of things-assisted two stream neural
Internet of Things Journal, 2021. network for anomaly detection in surveillance big video
[142] H. Wang, Z. Wu, H. Jiang, Y. Huang, J. Wang, S. Kopru, data, Future Generation Computer Systems 129 (2022) 286–
and T. Xie, Groot: An event-graph-based approach for root 297.
cause analysis in industrial settings, in: Proceedings of [155] W. Xu, L. Huang, A. Fox, D. Patterson, and M. Jordan,
IEEE/ACM International Conference on Automated ‘‘Online system problem detection by mining patterns of
Software Engineering, 2021. console logs,” in: 2009 ninth IEEE international conference
[143] A. Protogerou, S. Papadopoulos, A. Drosou, D. Tzovaras, I. on data mining. IEEE, 2009, pp. 588–597.
Refanidis, A graph neural network method for distributed [156] D. Duplyakin, R. Ricci, A. Maricq, G. Wong, J. Duerig, E.
anomaly detection in iot, Evolving Systems 12 (1) (2021) Eide, L. Stoller, M. Hibler, D. Johnson, K. Webb et al., The
19–36. design and operation of fCloudLabg”, in: 2019 USENIX
[144] F. Fusco, B. Eck, R. Gormally, M. Purcell, and S. Tirupathi, annual technical conference (USENIX ATC 19), 2019, pp.
Knowledge-and data-driven services for energy systems 1–14.
using graph neural networks, in: Proceedings of IEEE [157] X. Zhou, C. Xu, B. Kimmons, Detecting tourism
destinations using scalable geospatial analysis based on

This article can be downloaded from here: www.ijaems.com 21

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

cloud computing platform, Computers, Environment and intelligent transportation system, IEEE Transactions on
Urban Systems 54 (2015) 144–153. Vehicular Technology 69 (9) (2020) 9497–9508.
[158] K. Vassakis, E. Petrakis, I. Kopanakis, J. Makridis, and G. [172] S. Yang, W. Ma, X. Pi, S. Qian, A deep learning approach
Mastorakis, Locationbased social network data for tourism to real-time parking occupancy prediction in transportation
destinations, in: Proceedings of Big data and innovation in networks incorporating multiple spatio-temporal data
tourism, travel, and hospitality, 2019, pp. 105–114. sources, Transportation Research Part C: Emerging
[159] W. Wang, N. Kumar, J. Chen, Z. Gong, X. Kong, W. Wei, Technologies 107 (2019) 248–265.
H. Gao, Realizing the potential of the internet of things for [173] A. Paranjothi, M.S. Khan, S. Zeadally, A survey on
smart tourism with 5g and ai, IEEE Network 34 (6) (2020) congestion detection and control in connected vehicles, Ad
295–301. Hoc Networks 108 (2020).
[160] S.Y. Park, B. Pan, Identifying the next non-stop flying [174] C. Chen, B. Liu, S. Wan, P. Qiao, Q. Pei, An edge traffic
market with a big data approach, Tourism Management 66 flow detection scheme based on deep learning in an
(2018) 411–421. intelligent transportation system, IEEE Transactions on
[161] A. Kontogianni, E. Alepis, C. Patsakis, Promoting smart Intelligent Transportation Systems 22 (3) (2020) 1840–
tourism personalized services via a combination of deep 1852.
learning techniques, Expert Systems with Applications vol. [175] M.M. Iqbal, M.T. Mehmood, S. Jabbar, S. Khalid, A.
187, paper number 115964 (2022). Ahmad, G. Jeon, An enhanced framework for multimedia
[162] M. Mariani, R. Baggio, Big data and analytics in hospitality data: Green transmission and portrayal for smart traffic
and tourism: a systematic literature review, International system, Computers & Electrical Engineering 67 (2018)
Journal of Contemporary Hospitality Management no. 1, 291–308.
paper number 232 (2021). [176] R.S. Thejaswini and S. Rajaraajeswari, A real-time traffic
[163] J.C. Cepeda-Pacheco, M.C. Domingo, Deep learning and congestion avoidance framework for smarter cities, in:
internet of things for tourist attraction recommendations in Proceedings of AIP Conference Proceedings, vol. 2039, no.
smart cities, Neural Computing and Applications (2022) 1– 1, paper number 020009, 2018.
19. [177] S. Sarker, B. Wankum, T. Perey, M.M. Mau, J. Shimizu, R.
[164] J. Guerra-Montenegro, J. Sanchez-Medina, I. La na, D. Jones, B.S. Terry, A novel capsule-delivered enteric drug-
Sanchez-Rodriguez, I. Alonso-Gonzalez, and J. Del Ser, injection device for delivery of systemic biologics: A pilot
‘‘Computational intelligence in the hospitality industry: A study in a porcine model, IEEE Transactions on Biomedical
systematic literature review and a prospect of challenges,” Engineering 69 (6) (2022) 1870–1879.
Applied Soft Computing, vol. 102, p. 107082, 2021. [178] S. Sangodoyin, E.M. Ugurlu, M. Dey, M. Prvulovic, A.
[165] F. Piccialli, F. Giampaolo, G. Casolla, V.S. Di Cola, K. Li, Zajic´, Leveraging on-chip transistor switching for
A deep learning approach for path prediction in a location- communication and sensing in neural implants and
based iot system, Pervasive and Mobile Computing vol. 66, gastrointestinal devices, IEEE Transactions on Biomedical
paper number 101210 (2020). Engineering 69 (1) (2021) 377–389.
[166] Y.-C. Chang, C.-H. Ku, C.-H. Chen, Using deep learning [179] S. Li, P. Si, Z. Zhang, J. Zhu, X. He, N. Zhang, Dfca-net:
and visual analytics to explore hotel reviews and responses, Dual feature context aggregation network for bleeding areas
Tourism Management 80 (2020). segmentation in wireless capsule endoscopy images, Journal
[167] G. Díaz, H. Macià, V. Valero, J. Boubeta-Puig, F. Cuartero, of Medical and Biological Engineering 42 (2) (2022) 179–
An intelligent transportation system to control air pollution 188.
and road traffic in cities integrating cep and colored petri [180] M.R. Basar, F. Malek, K.M. Juni, M.S. Idris, M.I.M. Saleh,
nets, Neural Computing and Applications 32 (2) (2020) Ingestible wireless capsule technology: A review of
405–426. development and future indication, International Journal of
[168] P. Arthurs, L. Gillam, P. Krause, N. Wang, K. Halder, A. Antennas and Propagation vol. 2012, paper number 807165
Mouzakitis, A taxonomy and survey of edge cloud (2012).
computing for intelligent transportation systems and [181] I. Mehmood, M. Sajjad, S.W. Baik, Mobile-cloud assisted
connected vehicles, IEEE Transactions on Intelligent video summarization framework for efficient management
Transportation Systems (2021). of remote sensing data generated by wireless capsule
[169] Q. Cai, M. Abdel-Aty, Y. Sun, J. Lee, J. Yuan, Applying a sensors, Sensors 14 (9) (2014) 17 112–17 145.
deep learning approach for transportation safety planning by [182] R. Hamza, K. Muhammad, Z. Lv, F. Titouna, Secure video
using high-resolution transportation and land use data, summarization framework for personalized wireless capsule
Transportation Research Part A: Policy and Practice 127 endoscopy, Pervasive and Mobile Computing 41 (2017)
(2019) 71–85. 436–450.
[170] J. Wang, R. Chen, Z. He, Traffic speed prediction for urban [183] K. Muhammad, S. Khan, N. Kumar, J. Del Ser, S. Mirjalili,
transportation network: A path based deep learning Vision-based personalized wireless capsule endoscopy for
approach, Transportation Research Part C: Emerging smart healthcare: Taxonomy, literature review,
Technologies 100 (2019) 372–385. opportunities and challenges, Future Generation Computer
[171] X. Wang, J. Liu, T. Qiu, C. Mu, C. Chen, P. Zhou, A real- Systems 113 (2020) 266–280.
time collision prediction mechanism with deep learning for

This article can be downloaded from here: www.ijaems.com 22

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

[184] K. Muhammad, M. Sajjad, M.Y. Lee, S.W. Baik, Efficient [198] J.-C. Hsieh, A.-H. Li, C.-C. Yang, Mobile, cloud, and big
visual attention driven framework for key frames extraction data computing: contributions, challenges, and new
from hysteroscopy videos, Biomedical Signal Processing directions in tele cardiology, International Journal of
and Control 33 (2017) 161–168. Environmental Research and Public Health 10 (11) (2013)
[185] K. Muhammad, J. Ahmad, M. Sajjad, S.W. Baik, Visual 6131–6153.
saliency models for summarization of diagnostic [199] G. Fortino, D. Parisi, V. Pirrone, G. Di Fatta, Bodycloud: A
hysteroscopy videos in healthcare systems, SpringerPlus SaaS approach for community body sensor networks, Future
vol. 5, paper number 1495 (2016). Generation Computer Systems 35 (2014) 62–79.
[186] R. Shrestha, S.K. Mohammed, M.M. Hasan, X. Zhang, K.A. [200] C. Xia, J. Zhao, H. Cui, X. Feng, J. Xue, Dnntune:
Wahid, Automated adaptive brightness in wireless capsule Automatic benchmarking dnn models for mobile-cloud
endoscopy using image segmentation and sigmoid function, computing, ACM Transactions on Architecture and Code
IEEE Transactions on Biomedical Circuits and Systems 10 Optimization (TACO) 16 (4) (2019) 1–26.
(4) (2016) 884–892. [201] B. Kumar, G. Pandey, B. Lohani, S.C. Misra, A multi-
[187] J.-Y. He, X. Wu, Y.-G. Jiang, Q. Peng, R. Jain, Hookworm faceted cnn architecture for automatic classification of
detection in wireless capsule endoscopy images with deep mobile lidar data and an algorithm to reproduce point cloud
learning, IEEE Transactions on Image Processing 27 (5) samples for enhanced training, ISPRS journal of
(2018) 2379–2392. photogrammetry and remote sensing 147 (2019) 80–89.
[188] X. Wu, H. Chen, T. Gan, J. Chen, C.-W. Ngo, Q. Peng, [202] A.E. Eshratifar, M.S. Abrishami, M. Pedram, JointDNN: An
Automatic hookworm detection in wireless capsule efficient training and inference engine for intelligent mobile
endoscopy images, IEEE Transactions on Medical Imaging cloud computing services, IEEE Transactions on Mobile
35 (7) (2016) 1741–1752. Computing 20 (2) (2019) 565–576.
[189] Y. Yuan, J. Wang, B. Li, M.Q.-H. Meng, Saliency based [203] A. Katal, S. Dahiya, T. Choudhury, A survey on cloud
ulcer detection for wireless capsule endoscopy diagnosis, computing in energy management of the smart grids, Cluster
IEEE Transactions on Medical Imaging 34 (10) (2015) Computing (2021).
2046–2057. [204] W. Pedrycz, Towards green machine learning: challenges,
[190] I. Mehmood, M. Sajjad, S.W. Baik, Video summarization opportunities, and developments, Journal of Smart
based tele endoscopy: a service to efficiently manage visual Environments and Green Computing 2 (4) (2022) 163–174.
data generated during wireless capsule endoscopy [205] R. Schwartz, J. Dodge, N.A. Smith, O. Etzioni, Green ai,
procedure, Journal of medical systems 38 (9) (2014) 1–9. Communications of the ACM 63 (12) (2020) 54–63.
[191] Y. Yuan, M.Q.-H. Meng, Deep learning for polyp [206] H. Jin, Q. Song, and X. Hu, Auto-keras: An efficient neural
recognition in wireless capsule endoscopy images, Medical architecture search system, in: ACM SIGKDD Conference
physics 44 (4) (2017) 1379–1389. on Knowledge Discovery and Data Mining, 2019, pp. 1946–
[192] X. Jia, M.Q.-H. Meng, A deep convolutional neural network 1956.
for bleeding detection in wireless capsule endoscopy [207] A. Li, O. Spyra, S. Perel, V. Dalibard, M. Jaderberg, C.J.
images, 38th annual international conference of the IEEE Gu, D. Budden, T. Harley, and P. Gupta, A generalized
engineering in medicine and biology society (EMBC), IEEE framework for population based training, in: Proceedings of
2016 (2016) 639–642. ACM SIGKDD Conference on Knowledge Discovery and
[193] S. Jain, A. Seal, A. Ojha, A. Yazidi, J. Bures, I. Tacheci, O. Data Mining, 2019.
Krejcar, A deep cnn model for anomaly detection and [208] T. Elsken, J.H. Metzen, F. Hutter, Neural architecture
localization in wireless capsule endoscopy images, search: A survey, The Journal of Machine Learning
Computers in Biology and Medicine 137 (2021). Research 20 (1) (2019) 1997–2017.
[194] T. Aoki, A. Yamada, K. Aoyama, H. Saito, A. Tsuboi, A. [209] P. Ren, Y. Xiao, X. Chang, P.-Y. Huang, Z. Li, X. Chen, X.
Nakada, R. Niikura, M. Fujishiro, S. Oka, S. Ishihara, et al., Wang, A comprehensive survey of neural architecture
Automatic detection of erosions and ulcerations in wireless search: Challenges and solutions, ACM Computing Surveys
capsule endoscopy images based on a deep convolutional (CSUR) 54 (4) (2021) 1–34.
neural network, Gastrointestinal endoscopy 89 (2) (2019) [210] H. Li, P. Wu, Z. Wang, J. f. Mao, F.E. Alsaadi, and Z.N. Y,
357–363. ‘‘A generalized framework of feature learning enhanced
[195] X. Gu, K. Nahrstedt, A. Messer, I. Greenberg, D. Milojicic, convolutional neural network for pathology-image-oriented
Adaptive offloading for pervasive computing, IEEE cancer,” Computers in Biology and Medicine, vol. 151A,
Pervasive Computing 3 (3) (2004) 66–73. 2022, 106265.
[196] K. Yang, S. Ou, H.-H. Chen, On effective offloading [211] H. Li, N.Y. Zeng, P.S. Wu, K. Clawson, Cov-net: A
services for resource constrained mobile devices running computer-aided diagnosis method for recognizing covid-19
heavier mobile internet applications, IEEE Communications from chest x-ray images via machine vision, Expert Systems
Magazine 46 (1) (2008) 56–63. with Applications 207 (2022).
[197] A.P. Miettinen and J.K. Nurminen, Energy efficiency of [212] P. Wu, Z. Wang, B. Zheng, H. Li, F.E. Alsaadi, N. Zeng,
mobile clients in cloud computing, in: Proceedings of 2nd Aggn: Attention-based glioma grading network with multi-
USENIX Workshop on Hot Topics in Cloud Computing, scale feature extraction and multi-modal information fusion,
2010. Computers in Biology and Medicine 152 (2023).

This article can be downloaded from here: www.ijaems.com 23

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

[213] I. Makdoom, M. Abolhasan, H. Abbas, W. No, Blockchains learning, IEEE Transactions on Industrial Informatics 19 (1)
adoption in iot: the challenges, and a way forward, Journal (2022) 1080–1087.
of Network and Computer Applications 125 (2019) 251– [227] Z. Liu, Z. Gao, J. Wang, Q. Liu, J. Wei, et al., Ppefl: An edge
279. federated learning architecture with privacy-preserving
[214] H. Hu, Z. Salcic, L. Sun, G. Dobbie, P.S. Yu, X. Zhang, mechanism, Wireless Communications and Mobile
Membership inference attacks on machine learning: A Computing 2022 (2022).
survey, ACM Computing Surveys (CSUR) 54 (11s) (2022) [228] J. Fiosina, ‘‘Interpretable privacy-preserving collaborative
1–37. deep learning for taxi trip duration forecasting,” in Smart
[215] M. Jegorova, C. Kaul, C. Mayor, A.Q. O’Neil, A. Weir, R. Cities, Green Technologies, and Intelligent Transport
Murray-Smith, S.A. Tsaftaris, Survey: Leakage and privacy Systems: 10th International Conference, SMARTGREENS
at inference time, IEEE Transactions on Pattern Analysis 2021, and 7th International Conference, VEHITS 2021,
and Machine Intelligence (2022). Virtual Event, April 28–30, 2021, Revised Selected Papers.
[216] M. Abadi, A. Chu, I. Goodfellow, H.B. McMahan, I. Springer, 2022, pp. 392–411.
Mironov, K. Talwar, and L. Zhang, Deep learning with [229] B. Jan, H. Farman, M. Khan, M. Imran, I.U. Islam, A.
differential privacy, in: Proceedings of the 2016 ACM Ahmad, S. Ali, G. Jeon, Deep learning in big data analytics:
SIGSAC conference on computer and communications a comparative study, Computers & Electrical Engineering
security, 2016, pp. 308–318. 75 (2019) 275–287.
[217] C. Dwork, ‘‘Differential privacy: A survey of results,” in [230] J. Dean, G.S. Corrado, R. Monga, K. Chen, M. Devin, Q.V.
Theory and Applications of Models of Computation: 5th Le, M.Z. Mao, M. Ranzato, A. Senior, P. Tucker et al.,
International Conference, TAMC 2008, Xi’an, China, April Large scale distributed deep networks, in: Proceedings of
25-29, 2008. Proceedings 5. Springer, 2008, pp. 1–19. Advances in Neural Information Processing Systems, 2012.
[218] S. Meftah, B.H.M. Tan, C.F. Mun, K.M.M. Aung, B. [231] B. Abu-Salih, P. Wongthongtham, D. Zhu, K.Y. Chan, A.
Veeravalli, V. Chandrasekhar, Doren: toward efficient deep Rudra, Social Big Data Analytics: Practices, Techniques,
convolutional neural networks with fully homomorphic and Applications, Springer Nature, 2021.
encryption, IEEE Transactions on Information Forensics [232] A.A. Taha, W. Ramo, H.H.K. Alkhaffaf, Impact of external
and Security 16 (2021) 3740–3752. auditor–cloud specialist engagement on cloud auditing
[219] Z. Zhong, W. Bao, J. Wang, X. Zhu, X. Zhang, Flee: A challenges, Journal of Accounting & Organizational Change
hierarchical federated learning framework for distributed 17 (3) (2021) 309–331.
deep neural network over cloud, edge, and end device, ACM [233] Z. Zhang, C. Wu, D.W. Cheung, A survey on cloud
Transactions on Intelligent Systems and Technology (TIST) interoperability: taxonomies, standards, and practice, ACM
13 (5) (2022) 1–24. SIGMETRICS Performance Evaluation Review 40 (4)
[220] R. Gupta, I. Gupta, D. Saxena, A.K. Singh, A differential (2013) 13–22.
approach and deep neural network based data privacy- [234] H. Schulze, Cloud security report, Fortinet, Report, 2021.
preserving model in cloud environment, Journal of Ambient [Online]. Available:
Intelligence and Humanized Computing (2022) 1–16. https://www.fortinet.com/content/dam/fortinet/assets/analy
[221] J. Vasa, A. Thakkar, Deep learning: Differential privacy st-reports/arcybersecurity-cloud-security.pdf.
preservation in the era of big data, Journal of Computer [235] C. Ramalingam, P. Mohan, Addressing semantics standards
Information Systems (2022) 1–24. for cloud portability and interoperability in multi cloud
[222] M.H. Rahman, M.M. Mowla, S. Shanto, Differential privacy environment, Symmetry 13 (2) (2021) 317.
enabled deep neural networks for wireless resource [236] Available: https://insightaas.com/cloud-interoperability-
management, Mobile Networks and Applications 27 (5) andportability-necessary-or-nice-to-have/.
(2022) 2153–2162. [237] R. Ranjan, The cloud interoperability challenge, IEEE Cloud
[223] J. Xiong, H. Zhu, Real-time trajectory privacy protection Computing 1 (2) (2014) 20–24.
based on improved differential privacy method and deep [238] A. Romasanta and J. Wareham, Fair data through a federated
learning model, Journal of Cloud Computing 11 (1) (2022) cloud infrastructure: Exploring the science mesh, in:
1–15. Proceedings of European Conference on Information
[224] F. Gava, L.M. Bayati, A scalable algorithm for Systems, 2021.
homomorphic computing on multi-core clusters, 2022 21st [239] B. Abu-Salih, Domain-specific knowledge graphs: A
International Symposium on Parallel and Distributed survey, Journal of Network and Computer Applications 185
Computing (ISPDC), IEEE (2022) 57–64. (2021).
[225] S. Meftah, B.H.M. Tan, K.M.M. Aung, L. Yuxiao, L. Jie, B. [240] H. Mezni, M. Sellami, S. Aridhi, F.B. Charrada, Towards
Veeravalli, Towards high performance homomorphic big services: a synergy between service computing and
encryption for inference tasks on cpu: An mpi approach, parallel programming, Computing 103 (11) (2021) 2479–
Future Generation Computer Systems 134 (2022) 13–21. 2519.
[226] J.A. Alzubi, O.A. Alzubi, A. Singh, M. Ramachandran, [241] O. Adedugbe, E. Benkhelifa, R. Campion, F. Al-Obeidat,
Cloud-iiot-based electronic health record privacy- A.B. Hani, U. Jayawickrama, Leveraging cloud computing
preserving by cnn and blockchain-enabled federated for the semantic web: review and trends, Soft Computing 24
(8) (2020) 5999–6014.

This article can be downloaded from here: www.ijaems.com 24

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

[242] I. Grangel-González, F. Lösch, and A. ul Mehdi, Knowledge Workshop on Machine Learning in High-Performance
graphs for efficient integration and access of manufacturing Computing Environments, 2015, pp. 1–5.
data, in: Proceedings of 25th IEEE International Conference [258] R. Tanabe and A. Fukunaga, Success-history based
on Emerging Technologies and Factory Automation parameter adaptation for differential evolution, in:
(ETFA), vol. 1. IEEE, 2020, pp. 93–100. Proceedings of IEEE Congress on Evolutionary
[243] G. Ditzler, M. Roveri, C. Alippi, R. Polikar, Learning in Computation, 2013, pp. 71–78.
nonstationary environments: A survey, IEEE Computational [259] M. Wu, W. Su, L. Chen, Z. Liu, W. Cao, K. Hirota,
Intelligence Magazine 10 (4) (2015) 12–25. Weightadapted convolution neural network for facial
[244] J. Lu, A. Liu, F. Dong, F. Gu, J. Gama, G. Zhang, Learning expression recognition in human-robot interaction, IEEE
under concept drift: A review, IEEE transactions on Transactions on Systems, Man, and Cybernetics - Systems
knowledge and data engineering 31 (12) (2018) 2346–2363. 51 (3) (2021) 1473–1484.
[245] M. De Lange, R. Aljundi, M. Masana, S. Parisot, X. Jia, A. [260] M. Gong, J. Liu, A. Qin, K. Zhao, K.C. Tan, Evolving deep
Leonardis, G. Slabaugh, T. Tuytelaars, A continual learning neural networks via cooperative coevolution with
survey: Defying forgetting in classification tasks, IEEE backpropagation, IEEE Transactions on Neural Networks
transactions on pattern analysis and machine intelligence 44 and Learning Systems 32 (1) (2021) 420–434.
(7) (2021) 3366–3385. [261] M.R. Chen, B.P. Chen, G.Q. Zeng, K. Lu, P. Chu, An
[246] T. Khan, W. Tian, G. Zhou, S. Ilager, M. Gong, R. Buyya, adaptive fractional-order bp neural network based on
Machine learning (ml)–centric resource management in extremal optimization for handwritten digits recognition,
cloud computing: A review and future directions, Journal of Neuro computing 391 (2020) 260–272.
Network and Computer Applications (2022). [262] Z. Lu, G. Sreekumar, E. Goodman, W. Banzhaf, K. Deb,
[247] R.S. Sutton, A.G. Barto, Reinforcement learning: An V.N. Boddeti, Neural architecture transfer, IEEE
introduction, MIT press, 2018. Transactions on Pattern Analysis and Machine Intelligence
[248] V. François-Lavet, P. Henderson, R. Islam, M.G. Bellemare, 43 (2021).
and J. Pineau, An introduction to deep reinforcement [263] B. Ma, X. Li, Y. Xia, Y. Zhang, Autonomous deep learning:
learning, arXiv preprint arXiv:1811.12560, 2018. A genetic dcnn designer for image classification,
[249] Y. Li, Deep reinforcement learning: An overview, arXiv Neurocomputing 379 (2020) 152–161.
preprint arXiv:1701.07274, 2017. [264] F. Mattioli, D. Caetano, A. Cardoso, E. Naves, and E.
[250] B.R. Kiran, I. Sobh, V. Talpaert, P. Mannion, A.A. Al Lamounier, An experiment on the use of genetic algorithms
Sallab, S. Yogamani, P. Pérez, Deep reinforcement learning for topology selection in deep learning, Journal of Electrical
for autonomous driving: A survey, IEEE Transactions on and Computer Engineering, no. Article ID 3217542, 2019.
Intelligent Transportation Systems (2021) 4909–4926. [265] B. Wang, Y. Sun, B. Xue, and M. Zhang, A hybrid ga-pso
[251] F. Zeng, C. Wang, and S.S. Ge, ‘‘A survey on visual method for evolving architecture and short connections of
navigation for artificial agents with deep reinforcement deep convolutional neural networks, in: Proceedings of the
learning,” IEEE Access, vol. 8, pp. 135 426–135 442, 2020. 16th Pacific Rim International Conference on Artificial
[252] K. Arulkumaran, M.P. Deisenroth, M. Brundage, A.A. Intelligence, 2019.
Bharath, Deep reinforcement learning: A brief survey, IEEE [266] M. Suganuma, S. Shirakawa, and T. Nagao, A genetic
Signal Processing Magazine 34 (6) (2017) 26–38. programming approach to designing convolutional neural
[253] K. Shao, Z. Tang, Y. Zhu, N. Li, and D. Zhao, A survey of network architectures, in: Proceedings of Twenty-Seventh
deep reinforcement learning in video games, arXiv preprint International Joint Conference on Artificial Intelligence,
arXiv:1912.10944, 2019. 2017, pp. 5369–5373.
[254] A. Gupta, A. Anpalagan, L. Guan, A.S. Khwaja, Deep [267] A.D. Martinez, J. Del Ser, E. Villar-Rodriguez, E. Osaba, J.
learning for object detection and scene perception in self- Poyatos, S. Tabik, D. Molina, F. Herrera, Lights and
driving cars:survey, challenges, and open issues, Array vol. shadows in evolutionary deep learning: Taxonomy, critical
10, paper number 100057 (2021). methodological analysis, cases of study, learned lessons,
[255] N.Y. Zeng, P.S. Wu, Z.D. Wang, H. Li, W.B. Liu, X.H. Liu, recommendations and challenges, Information Fusion 67
A small-sized object detection oriented multi-scale feature (2021) 161–194.
fusion approach with application to defect detection, IEEE [268] T. Serizawa and H. Fujita. (2021) Optimization of
Transactions on Instrumentation and Measurement 71 convolutional neural network using the linearly decreasing
(2022). weight particle swarm optimization. [Online]. Available:
[256] N. Rodríguez-Barroso, A.R. Moya, J.A. Fernández, E. https://arxiv.org/abs/2001.05670.
Romero, E. Martínez-Cámara, and F. Herrera, Deep [269] J. Fregoso, C.I. Gonzalez, G.E. Martinez, Optimization of
learning hyper-parameter tuning for sentiment analysis in convolutional neural networks architectures using pso for
twitter based on evolutionary algorithms, in: 2019 Federated sign language recognition, Axioms vol. 10, number paper
Conference on Computer Science and Information Systems 139 (2021).
(FedCSIS). IEEE, 2019, pp. 255–264. [270] Z. Fouad, M. Alfonse, M. Roushdy, A.-B.M. Salem, Hyper-
[257] S.R. Young, D.C. Rose, T.P. Karnowski, S.-H. Lim, and parameter optimization of convolutional neural network
R.M. Patton, Optimizing deep learning hyper-parameters based on particle swarm optimization algorithm, Bulletin of
through an evolutionary algorithm, in: Proceedings of the Electrical Engineering and Informatics 10 (6) (2021) pp.

This article can be downloaded from here: www.ijaems.com 25

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/
Awan International Journal of Advanced Engineering, Management and Science, 9(10) -2023

[271] X. Chen, Y. Sun, M. Zhang, D. Peng, Evolving deep

convolutional variational autoencoders for image
classification, IEEE Transactions on Evolutionary
Computation 25 (5) (2021).
[272] T. Zhang, C. Lei, Z. Zhang, X.-B. Meng, C.L.P. Chen, As-
nas: Adaptive scalable neural architecture search with
reinforced evolutionary algorithm for deep learning, IEEE
Transactions on Evolutionary Computation 25 (5) (2021).
[273] L. Xie and A. Yuille, Genetic cnn, in: Proceedings of IEEE
International Conference on Computer Vision, 2017.
[274] S. Pimminger, S. Wagner, W. Kurschl, and J. Heinzelreiter,
Optimization as a service: On the use of cloud computing
for metaheuristic optimization, in: Proceedings of the
International Conference on Computer Aided Systems
Theory, 2013, p. 348–355.
[275] A.D. Martinez, J.D. Ser, E. Villar-Rodriguez, E. Osaba, J.
Poyatos, S. Tabik, D. Molina, F. Herrera, Lights and
shadows in evolutionary deep learning taxonomy, critical
methodological analysis, cases of study, learned lessons,
recommendations and challenges, Information Fusion 67
(2021) 161–194.

This article can be downloaded from here: www.ijaems.com 26

©2023 The Author(s). Published by Infogain Publication, This work is licensed under a Creative Commons Attribution 4.0 License.
http://creativecommons.org/licenses/by/4.0/