Name: Abodh kumar

SID: 75014
Sub: AML

Submitted by: Submitted to:

Abodh Kumar Mrs. Priyanka Gupta
(Sid-75014 )
Q1. Discuss Hilbert-Huang transform and Deep Convolutional
Neural Network for Bearing Fault Classification.
Ans:- The Hilbert-Huang Transform (HHT) represents a desperate attempt to
break the suffocating hold on the field of data analysis by the twin assumptions
of linearity and stationarity. Unlike spectrograms, wavelet analysis, or the
Wigner-Ville Distribution, HHT is truly a time-frequency analysis, but it does
not require an a priori functional basis and, therefore, the convolution
computation of frequency. The method provides a magnifying glass to examine
the data, and also offers a different view of data from nonlinear processes,
with the results no longer shackled by spurious harmonics — the artifacts of
imposing a linearity property on a nonlinear system or of limiting by the
uncertainty principle, and a consequence of Fourier transform pairs in data
analysis. This is the first HHT book containing papers covering a wide variety of
interests. The chapters are divided into mathematical aspects and applications,
with the applications further grouped into geophysics, structural safety and
visualization.

Deep Convolutional Neural Network for Bearing Fault Classification.

With the advent of the 4th industrial revolution, industries across the globe are
using artificial intelligence (AI) to improve their processes and increase
efficiency to meet the ever-rising customer demands. In this rapidly changing
landscape of technology, organizations across the globe having increased the
presence of sensors on the production floor with the motivation of gathering
data that can give them valuable insights into their processes [1]. This sensory
data contains rich information about the machine and its effective analyses
using AI can contribute significantly towards preventive maintenance, quality
control, and increased process efficiency [2]. Realizing these obvious benefits
of cost-effective tools like AI, organizations across the world are turning
towards smarter technologies.

Driven by the keenness of the industry to embrace advanced digital

technologies, many researchers are using different signal processing methods
and coupling them with machine learning algorithms to address some of the
complicated research problems. For example, Thomazella et al. used digital
signal processing techniques such as short-time Fourier transform (STFT) and
the ratio of power (ROP) to extract features from vibrations signals captured to
monitor chatter phenomenon during the grinding process [3]. In another paper
on signal processing on vibration data, Zoltan et al., have demonstrated that
signal processing techniques such as Discrete Wavelet and Wavelet Packet
Transform are effective in extracting features from the frequency domain for
fault detection [4]. Their simulated results proved the techniques are even
capable of predicting abnormalities exploring long-term tendencies of the
detected signals. Signal processing techniques have been performed on
acoustic signals as well. Adam et al. have proposed a signal processing
technique named MSAF-RATIO-24-MULTIEXPANDED-FILTER-8. This technique
is used on acoustic signals captured from electric motors and extracted
features are used to classify motor faults [5].

Q2. Write Type of government problems appropriate for AI

applications.
Ans- Type of government problems appropriate for AI applications.

Resource Allocation:
 Administrative support is needed to speed up task completion
 Inquiry response times are long due to insufficient support

Large Datasets:
 Dataset is too large for employees to work with efficiently
 Internal and external datasets can be combined to enhance outputs and
insights
 Data is highly structured with years of history

Experts Shortage:
 Basic questions can be answered, freeing up time for experts
 Niche issues can be learned to support experts in research
Predictable Scenario:
 Situation is predictable based on historical data
 Prediction will help with time-sensitive responses

Procedural:
 Task is repetitive in nature
 Inputs/outputs have binary answer

Diverse Data:
 Data includes visual/spatial and auditory/linguistic information
 Qualitative and quantitative data needs to be summarized regularly

Q3. Explain Chat bots for communication between citizen &

government and Chat bots in the public sector.
Ans:-Artificial Intelligence (AI) techniques have been extensively used to
support and enhance the quality of decision making and problem solving in
different industries for many years. This is through the exploitation of various
types of machine intelligence, including natural language comprehension,
robotics, expert systems, neural networks and machine learning (Buzzle, n.d.;
William, Schatsky, & Viechnicki, 2017). Admittedly, the most rising AI trend in
the private sector nowadays concerns the utilization of applications that
interact with users in a conversational format and mimic human conversation,
known as “conversational agents”, “chatbots” or simply “bots” (Abu Shawar &
Atwell, 2007; Klopfenstein, Delpriori, Malatini, & Bogliolo, 2017; Poola, 2017).
Chatbots are intelligent agents (defined as devices that perceive their
environment and take actions that maximize their chance of success at some
goal (Russell & Norvig, 2009), which have the ability to understand a spoken
language and use speech communication as user interface. They consist of a
specific AI-based software category developed by companies to automate
communications and management of transactions with their customers.
Indicative applications of chatbots in the private sector include the
implementation of virtual assistants (e.g. Siri, Alexa, Google now, Cortana) in
various sectors, such as banking (implementing transactions), media (news
provision), tourism (booking hotels or tickets), retail, stock market, insurance,
gaming agencies, telecommunications, etc. (Business Insider, 2017; Dole,
Sansare, Harekar, & Athalye, 2015; Lokot & Diakopoulos, 2016; Poola, 2017;
Zsarnoczky, 2017). The first generation of chatbots has limited capabilities
allowing them to respond to simple rule-based queries; however, due to
recent AI advancements and the abundance of available data, chatbots can
now perform more complex tasks and even complete proactive transactions.

Motivated by the efficiency gains reported by private sector firms, AI

technologies have also started being adopted by government agencies to take
on significantly complex tasks in diverse domains, e.g. health, social welfare,
public safety, taxation, and education (Business Insider, 2016; Capgemini,
2017; Kerly, Hall, & Bull, 2007; Mahdavi & Shepherd, 2004). A recent report by
Harvard identifies 6 types of government problems, for which AI applications
are considered as highly appropriate: resource allocation, large datasets,
experts shortage, predictable scenarios, procedural and repetitive tasks,
diverse data aggregation and summarisation (Mehr, 2017). A lot of benefits of
ΑΙ applications in the public sector are broadly reported, namely cost savings,
alleviation of public servants' workload, increase of productivity, creation of
new employment opportunities, solution of resource allocation problems,
public service delivery and improvement of citizens' satisfaction (through
personalisation and 24/7 availability) (Capgemini, 2017; Eggers. William et al.,
2017; White House, 2016). In particular, chatbots and other types of AI
solutions (e.g. machine learning algorithms, process automation and image
recognition software) can significantly reduce the administrative burden of
public organizations and advance the communication between government
and citizens within the provision of public services, which has been
problematic for a long time.

Despite these initial benefits, the adoption of chatbots in the public sector
delineates a new research domain posing a series of challenges regarding the
exploitation of AI capabilities and their contribution in public services delivery.
For instance, a well-known challenge in developing AI applications concerns
the extraction and representation of the expertise needed to develop their
knowledge base. This challenge gets higher in the public sector, where both
governmental experts' knowledge and a multitude of data (legislative,
operational) have to be codified into a format that can be machine-exploitable.
Data quality has also to be ensured through efficient management, structuring
and aggregation of diverse data. Schoemaker and Tetlock (2017) suggest a
hybrid approach, combining both humans and computers to reach the so-
called ‘superior judgements’, as a means to address the lack of broader and
contextual intelligence of humans in AI solutions. In addition, it should be
noted that a number of ethical and social barriers impede the adoption of AI
technologies in the public sector, mainly rooted in the opposition on the
replacement of employees by machines and the lack of citizens' trust on
machine intelligence (Capgemini, 2017; Centre for Public Impact, 2017).

Chat bots in the public sector.

hatbots are becoming an integral part of our society, as tech companies invest
significant resources in developing and implementing the technology to
improve user experience. In simple terms, chatbots are computer programs
that leverage machine learning and artificial intelligence (AI) to complete tasks
while mimicking human conversation.

Public agencies are developing their own chatbots to transform their service
delivery. First, agencies are reducing employees’ workload and response times
by delegating mundane and routine tasks to chatbots, saving human labor for
more technical and nuanced tasks. For instance, North Carolina’s Innovation
Center (iCenter) is piloting chatbots to free up internal IT help desk personnel
from answering mundane queries (e.g. password resetting). The iCenter found
that 80 to 90 percent of queries submitted to the IT help desk involve
recovering account information such as passwords and user IDs. By leveraging
AI to help with the routine requests, IT personnel can focus on complex issues
that require human assistance.

Q4.Describe applications of ML and artificial intelligence in

transportation, logistics , Incident detection. and Limitations
of AI techniques in transportation.
Ans:-

Machine learning, as a subset of the tools of artificial intelligence, is a broad

area of computer science that provides machines and autonomous systems
with the ability to learn and improve on previous experience. In other words,
machine learning focuses on the development of computer programs that can
autonomously access data and use it for learning and self-improvement. This
core objective of machine learning can be achieved through many types of
learning, including supervised learning, unsupervised learning, reinforcement
learning, self-learning, feature learning, and association rules, among others.
Various models that implement these machine learning types have been used
and researched for machine learning systems. The best models are known to
include architectures such as artificial neural networks, support vector
machines, decision trees, evolutionary algorithms, Petri nets, and regression
algorithms. It seems that machine learning is related to, and overlaps with,
other fields of information theory, e.g., optimisation, statistics, and data
mining.

The issues in transportation are becoming a challenge due to the increased

globalisation of the economy, population growth, development of industrial
production, safety concerns, and environmental degradation. Traditional
transportation systems are pushed to the limit, and due to the ever-increasing
traffic volume, urban development often reaches a bottleneck. With the
current availability of computational power and massive progress in the theory
of artificial intelligence, machine learning models seem to be a suitable tool to
overcome the previously mentioned challenges. In some cases, such as logistic
service planning, event and object detection from surveillance, video, or
analysis of traveller’s behaviour, machine learning-based solutions have
already proved themselves to have higher performance compared to
conventional solutions. Other possible applications are still the object of
research and development.

This Special Issue is focused on the reporting of new and innovative

applications of machine learning methods to solve transportation problems.
Submissions, which provide reliable, robust, and verified approaches, are
especially welcome. Among others, researchers and academics, who intensely
cooperate with industrial and transportation companies, are encouraged to
publish their contributions. We welcome both original research and review
articles.

Potential topics include but are not limited to the following:

  Traffic analysis and control, including monitoring, public transportation
management, etc.
 Decision support for intelligent transportation systems
 Modelling of aspects of transportation systems including user behaviour,
travel demand, and infrastructure usage
 Applications of modelling to traffic analysis, control, and optimisation
 Smart city logistics

from one language to another, a self-driving car, a product recommendation

system on an ecommerce website such as amazon.com and so on. Considering
these diverse applicationscenarios, machine learning has become a popular
method of developing computer systems inseveral domains in the past decade,
and applied linguistics is no exception.This entry explores the relevance of
machine learning in the context of technological applications for language
learning. Starting with a brief introduction to machine learning, thearticle will
focus on the relevance, methods, and uses of machine learning in three broad
application areas in language learning and applied linguistics: assessment,
support, and analytics.

Due to the tremendous population growth in the country, the use of vehicles
and other transportation means have increased which has led to traffic
congestion and road accidents. Hence, there is a demand for intelligent
transportation systems in the country that can provide safe and reliable
transportation while maintaining environmental conditions such as pollution,
CO2 emission, and energy consumption. This paper focuses on providing an
overview and applications of how Artificial intelligence (AI) and Machine
Learning (ML) can be applied to develop an Intelligent Transportation system
that can address the issues of traffic congestion and road safety to prevent
accidents. We will then re-view various ML approaches to detect road
anomalies for avoiding obstacles, predict re-al-time traffic flow to achieve
smart and efficient transportation, detect and pre-vent road accidents to
ensure safety, using smart city lights to save energy and smart infrastructure to
achieve efficient transportation. Next, we review various AI approaches such as
safety and emergency management system to provide safety to the public,
autonomous vehicles to provide economical and reliable transportation. We
then propose smart parking management and how it can be used to find
parking spaces or spots conveniently, incident detection which detects the
traffic incidents or accidents real-time provides a report. Finally, we conclude
with predictive models and how the algorithms utilize sensor data to develop
an Intelligent Transportation System

Q5. What is Demand forecasting? And also explain its methods and
Predictor variables in demand forecasting.
Ans:- It is a technique for estimation of probable demand for a product or
services in the future. It is based on the analysis of past demand for that
product or service in the present market condition. Demand forecasting should
be done on a scientific basis and facts and events related to forecasting should
be considered.

Therefore, in simple words, we can say that after gathering information about
various aspect of the market and demand based on the past, an attempt may
be made to estimate future demand. This concept is called forecasting of
demand.

For example, suppose we sold 200, 250, 300 units of product X in the month of
January, February, and March respectively. Now we can say that there will be a
demand for 250 units approx. of product X in the month of April, if the market
condition remains the same.

Methods of Demand Forecasting

There is no easy or simple formula to forecast the demand. Proper judgment
along with the scientific formula is needed to correctly predict the future
demand for a product or service. Some methods of demand forecasting are
discussed below:

1] Survey of Buyer’s Choice

When the demand needs to be forecasted in the short run, say a year, then the
most feasible method is to ask the customers directly that what are they
intending to buy in the forthcoming time period. Thus, under this method,
potential customers are directly interviewed. This survey can be done in any of
the following ways:

 Complete Enumeration Method: Under this method, nearly all the

potential buyers are asked about their future purchase plans.
  Sample Survey Method: Under this method, a sample of potential buyers
are chosen scientifically and only those chosen are interviewed.
 End-use Method: It is especially used for forecasting the demand of the
inputs. Under this method, the final users i.e. the consuming industries
and other sectors are identified. The desirable norms of consumption of
the product are fixed, the targeted output levels are estimated and
these norms are applied to forecast the future demand of the inputs.

Hence, it can be said that under this method the burden of demand forecasting
is on the buyer. However, the judgments of the buyers are not completely
reliable and so the seller should take decisions in the light of his judgment also.

The customer may misjudge their demands and may also change their
decisions in the future which in turn may mislead the survey. This method is
suitable when goods are supplied in bulk to industries but not in the case of
household customers.

2] Collective Opinion Method

Under this method, the salesperson of a firm predicts the estimated future
sales in their region. The individual estimates are aggregated to calculate the
total estimated future sales. These estimates are reviewed in the light of
factors like future changes in the selling price, product designs, changes in
competition, advertisement campaigns, the purchasing power of the
consumers, employment opportunities, population, etc.

The principle underlying this method is that as the salesmen are closest to the
consumers they are more likely to understand the changes in their needs and
demands. They can also easily find out the reasons behind the change in their
tastes.

Therefore, a firm having good sales personnel can utilize their experience to
predict the demands. Hence, this method is also known as Salesforce opinion
or Grassroots approach method. However, this method depends on the
personal opinions of the sales personnel and is not purely scientific.

3] Barometric Method
This method is based on the past demands of the product and tries to project
the past into the future. The economic indicators are used to predict the future
trends of the business. Based on future trends, the demand for the product is
forecasted. An index of economic indicators is formed. There are three types of
economic indicators, viz. leading indicators, lagging indicators, and coincidental
indicators.