Unit 6 :Artificial intelligence

What is Neural Network in Artificial Intelligence(ANN)?

ANN stands for Artificial Neural Networks. Basically, it’s a computational
model. That is based on structures and functions of biological neural networks.
Although, the structure of the ANN affected by a flow of information. Hence,
neural network changes were based on input and output.
Basically, we can consider ANN as nonlinear statistical data. That means
complex relationship defines between input and output. As a result, we found
different patterns. Also, we call the ANN as a neural network.
Structure of Artificial Neural Network
Generally, the working of a human brain by making the right connections is the
idea behind ANNs. That was limited to use of silicon and wires as living
neurons and dendrites.
Here, neurons, part of human brain. That was composed of 86 billion nerve
cells. Also, connected to other thousands of cells by Axons.Although, there are
various inputs from sensory organs. That was accepted by dendrites.
As a result, it creates electric impulses. That is used to travel through the
Artificial neural network. Thus, to handle the different issues, neuron send a
message to another neuron.

Basic Structure of Artificial Neural Network

As a result, we can say that ANNs are composed of multiple nodes. That imitate
biological neurons of the human brain. Although, we connect these neurons by
links. Also, they interact with each other.
Although, nodes are used to take input data. Further, perform simple operations
on the data. As a result, these operations are passed to other neurons. Also,
output at each node is called its activation or node value.
As each link is associated with weight. Also, they are capable of learning. That
takes place by altering weight values. Hence, the following illustration shows a
simple ANN −

Artificial Neural Network Structure

Artificial Neural Networks are a special type of machine learning algorithms
that are modeled after the human brain. That is, just like how the neurons in our
nervous system are able to learn from the past data, similarly, the ANN is able
to learn from the data and provide responses in the form of predictions or
classifications.

ANNs are nonlinear statistical models which display a complex relationship

between the inputs and outputs to discover a new pattern. A variety of tasks
such as image recognition, speech recognition, machine translation as well as
medical diagnosis makes use of these artificial neural networks.

An important advantage of ANN is the fact that it learns from the example data
sets. Most commonly usage of ANN is that of a random function
approximation. With these types of tools, one can have a cost-effective method
of arriving at the solutions that define the distribution. ANN is also capable of
taking sample data rather than the entire dataset to provide the output result.
With ANNs, one can enhance existing data analysis techniques owing to their
advanced predictive capabilities.

Don’t forget to check DataFlair’s leading blog on Recurrent Neural

Networks.
Artificial Neural Networks Architecture
The functioning of the Artificial Neural Networks is similar to the way neurons
work in our nervous system. The Neural Networks go back to the
early 1970s when Warren S McCulloch and Walter Pitts coined this term. In
order to understand the workings of ANNs, let us first understand how it is
structured.

In a neural network, there are three essential layers –

Input Layers
The input layer is the first layer of an ANN that receives the input information
in the form of various texts, numbers, audio files, image pixels, etc.

Hidden Layers
In the middle of the ANN model are the hidden layers. There can be a single
hidden layer, as in the case of a perceptron or multiple hidden layers. These
hidden layers perform various types of mathematical computation on the input
data and recognize the patterns that are part of.

Output Layer
In the output layer, we obtain the result that we obtain through rigorous
computations performed by the middle layer.
In a neural network, there are multiple parameters and hyperparameters that
affect the performance of the model. The output of ANNs is mostly dependent
on these parameters. Some of these parameters are weights, biases, learning
rate, batch size etc. Each node in the ANN has some weight.

Each node in the network has some weights assigned to it. A transfer function is
used for calculating the weighted sum of the inputs and the bias.

After the transfer function has calculated the sum, the activation function
obtains the result. Based on the output received, the activation functions fire the
appropriate result from the node. For example, if the output received is above
0.5, the activation function fires a 1 otherwise it remains 0.
Some of the popular activation functions used in Artificial Neural Networks are
Sigmoid, RELU, Softmax, tanh etc.
Based on the value that the node has fired, we obtain the final output. Then,
using the error functions, we calculate the discrepancies between the predicted
output and resulting output and adjust the weights of the neural network through
a process known as backpropagation.
ANNs are part of an emerging area in Machine Learning known as Deep
Learning.

.
Back Propagation in Artificial Neural Networks
In order to train a neural network, we provide it with examples of input-output
mappings. Finally, when the neural network completes the training, we test the
neural network where we do not provide it with these mappings. The neural
network predicts the output and we evaluate how correct the output is using the
various error functions. Finally, based on the result, the model adjusts the
weights of the neural networks to optimize the network following gradient
descent through the chain rule.

Types of Artificial Neural Networks

Generally, there are two types of ANN. Such as FeedForward and Feedback.
a. FeedForward ANN
In this network flow of information is unidirectional. A unit used to send
information to another unit that does not receive any information. Also, no
feedback loops are present in this. Although, used in recognition of a pattern. As
they contain fixed inputs and outputs.

Types of Artificial Neural Networks – FeedForward ANN

b. FeedBack ANN
In this particular Artificial Neural Network, it allows feedback loops. Also, used
in content addressable memories.

Types of Artificial Neural Networks – FeedBack ANN

How Does Artificial Neural Networks Works?
In this topology diagrams, you will learn everything in a detailed manner.
In this, each arrow represents a connection between two neurons. Also, they
used to indicate the pathway for the flow of information. As it was noticed that
each connection has a weight, an integer number. That used to controls the
signal between the two neurons.
If the output is good that was generated by the network then we don’t require to
adjust the weights. Although, if poor output generated. Then surely system will
alter the weight to improve results.

Machine Learning in ANNs

As there are too many Machine learning strategies are present, let’s see them
one by one:
a. Supervised Learning
Generally, in this learning a teacher is present to teach. That teacher must be
aware of ANN.
For example:
The teacher feeds only example data. That teacher already knows the answers.
b. Unsupervised Learning
If there is present no data set. Then we need this learning technique.
c. Reinforcement Learning
As this Machine learning technique is based on the observation. Although, if it’s
negative the networks need to adjust its weights. That is able to make a different
required decision the next time.
Back Propagation Algorithm
Generally, we use to call it as training and learning algorithm. As these
networks are ideal for simple Pattern Recognition and Mapping Task.

Bayesian Networks (BN)

Basically, we use to call it as graphical structures. Generally, we use this
network to represent probabilistic representation. This represents among a set of
random variables. Also, we used to call this network as Belief networks or
Bayes Nets.
In these networks, each node represents a random variable with specific
propositions.
In this only constraint arcs present in BN. Thus, doesn’t need to return node by
following directed arcs.
Hence, we can say BNs are known as Directed Acyclic Graphs (DAGs). Hence,
we use BNs to handle multivalued variables simultaneously.
Thus, BN variables composed of two dimensions −
 Range of prepositions
 Probability assigned to each of the prepositions.

Artificial Neural Networks Applications
Artificial Neural Network used to perform a various task. Also, this task
performs that are busy with humans but difficult for a machine.
a. Aerospace
Generally, we use ANN a for Autopilot aircrafts. They used for aircraft fault
detection.
b. Military
In various ways, we use ANN an in the military. Such as Weapon orientation
and steering, target tracking.
c. Electronics
Basically, we use an Artificial neural network in electronics in many ways. That
are code sequence prediction, IC chip layout, and chip failure analysis.
d. Medical
As medical has too many machines. That use in various ways.Such as cancer
cell analysis, EEG and ECG analysis.
e. Speech
We use ANN in speech recognition and speech classification.
f. Telecommunications
Generally, it has different applications. Thus, we use an Artificial neural
network in many ways. Such as image and data compression, automated
information services.
g. Transportation
Generally, we use an Artificial neural network in transportation in many ways.
That are truck Brake system diagnosis and vehicle scheduling, routing systems.
h. Software
It also uses an ANN in pattern Recognition. Such as in facial recognition,
optical character recognition, etc.
i. Time Series Prediction
We use an Artificial neural network to predict time. Also, we use ANNs to
make predictions on stocks and natural calamities.

ANN Advantages
1) It has a parallel processing ability. It has the numerical strength that performs
more than one task at the same time.
2) Failure of one element of the network does not affect the working of the
whole system. This characteristic makes it fault-tolerant.
3) A neural network learns from the experience and does not need
reprogramming.

Disadvantages of ANN
1) Its black-box nature is the most prominent disadvantage of ANN. The neural
network does not give the proper explanation of determining the output. It
reduces trust in the network.
2) The duration of the development of the network is unknown.
3) There is no assurance of proper network structure. There is no proper rule to
determine the structure.

Advantages of neural networks compared to conventional computers:

1. Neural networks have the ability to learn by themselves and produced the
output that is not limited to the input provided to them.
2. The input is stored in its own networks instead of the database. Hence, data
loss does not change the way it operates.
3. The neural network performs multiple tasks without affecting system
performance.
4. If any information is missing, the network can detect the fault and still
produce the output.
Due to these above advantages, a neural network has many applications as
compared with conventional computers such as image processing, route
detection, forecasting, speech recognition, language processing and translation.
Attributes of Neural Networks

With the human-like ability to problem-solve — and apply that skill to huge
datasets — neural networks possess the following powerful attributes:

 Adaptive Learning: Like humans, neural networks model non-linear and

complex relationships and build on previous knowledge. For example,
software uses adaptive learning to teach math and language arts.
 Self-Organization: The ability to cluster and classify vast amounts of
data makes neural networks uniquely suited for organizing the
complicated visual problems posed by medical image analysis.
 Real-Time Operation: Neural networks can (sometimes) provide real-
time answers, as is the case with self-driving cars and drone navigation.
 Prognosis: NN’s ability to predict based on models has a wide range of
applications, including for weather and traffic.
 Fault Tolerance: When significant parts of a network are lost or missing,
neural networks can fill in the blanks. This ability is especially useful in
space exploration, where the failure of electronic devices is always a
possibility.

Tasks Neural Networks Perform

Neural networks are highly valuable because they can carry out tasks to make
sense of data while retaining all their other attributes. Here are the critical tasks
that neural networks perform:

 Classification: NNs organize patterns or datasets into predefined classes.

 Prediction: They produce the expected output from given input.
 Clustering: They identify a unique feature of the data and classify it
without any knowledge of prior data.
 Associating: You can train neural networks to "remember" patterns.
When you show an unfamiliar version of a pattern, the network associates
it with the most comparable version in its memory and reverts to the
latter.

The Challenges of Neural Networks

As useful as neural networks can be, challenges in the field abound:

 Training: A common criticism of neural networks, particularly in

robotics applications, is that excessive training for real-world operations
is mandatory. One way to overcome that hurdle is by randomly shuffling
 training examples. Using a numerical optimization algorithm, small steps
— rather than large steps — are taken to follow an example. Another way
is by grouping examples in so-called mini-batches. Improving training
efficiencies and convergence capabilities is an ongoing research area for
computer scientists.
 Theoretical Issues: Unsolved problems remain, even for the most
sophisticated neural networks. For example, despite its best efforts,
Facebook still finds it impossible to identify all hate speech and
misinformation by using algorithms. The company employs thousands of
human reviewers to resolve the problem. In general, because computers
aren’t human, their ability to be genuinely creative — prove math
theorems, make moral choices, compose original music, or deeply
innovate — is beyond the scope of neural networks and AI.
 Inauthenticity: The theoretical challenges we address above arise
because neural networks don’t function exactly as human brains do —
they operate merely as a simulacrum of the human brain. The specifics of
how mammalian neurons code information is still an unknown. Artificial
neural networks don’t strictly replicate neural function, but rather use
biological neural networks as their inspiration. This process allows
statistical association, which is the basis of artificial neural networks. An
ANN’s learning process isn’t identical to that of a human, thus, its
inherent (at least for now) limitations.
 Hardware Issues: This century’s focus on neural networks is due to the
million-fold increase in computing power since 1991. More hardware
capacity has enabled greater multi-layering and subsequent deep learning,
and the use of parallel graphics processing units (GPUs) now reduces
training times from months to days. Despite the great strides of NNs in
very recent years, as deep neural networks mature, developers need
hardware innovations to meet increasing computational demands. The
search is on, and new devices and chips designed specifically for AI are
in development. A 2018 New York Times article, “Big Bets on A.I. Open
a New Frontier for Chip Startups, Too,” reported that “venture capitalists
invested more than $1.5 billion in chip startups” in 2017.
 Hybrids: A proposal to overcome some of the challenges of neural
networks combines NN with symbolic AI, or human-readable
representations of search, logic, and problems. To successfully duplicate
human intelligence, it’s vital to translate the procedural
knowledge or implicit knowledge ( the skills and knowledge not readily
accessible by conscious awareness) humans possess into an unequivocal
form that uses symbols and rules. So far, the difficulties of developing
symbolic AI have been unresolvable — but that status may soon change.
 Artificial Intelligence
John McCarthy was the first who defined the term Artificial Intelligence in
1956. He explained AI as:

“ The Science and Engineering of making Intelligent Machines. – John

McCarthy”

Artificial Intelligence Can Likewise Be defined because of the evolution Of

computer technologies that are designed for performing tasks which call for
human intelligence, like decision making, object discovery, fixing complex
troubles and so forth.

3. Types Of Learning In Artificial Intelligence

 Artificial Narrow Intelligence(ANI)

 Artificial General Intelligence(AGI)
 Artificial Super Intelligence(ASI)

These will be the three phases where AI may evolve, instead Compared to 3
types of Artificial Intelligence.

Let us understand each point in depth.

> Artificial Narrow Intelligence

Additionally known as Weak AI, ANI is your stage of Artificial Intelligence

between machines that could perform a narrowly defined set of specific tasks.
As of this phase, the machine doesn’t possess any believing capability, but it
simply plays a set of functions that are pre-defined.

Cases of Weak AI include Siri, Alexa, Self-driving automobiles, Alpha-Go,

Sophia the Ninja and so on. Just about all of the AI-based methods built till this
date fall below the category of Weak AI.
> Artificial General Intelligence

Additionally known as Strong AI, AGI is your stage at the development of

Synthetic Intelligence where machines will probably possess the ability to
believe and make conclusions like people human beings.

There are now no Present cases of Strong AI, yet, It is considered that we will
soon be able to generate machines which can be as smart as human beings.

Study AI Is Regarded as a hazard to human existence with lots of Scientists,

including Stephen Hawking who stated that:

“The development of complete Artificial Intelligence could stop the end of the
Human Era…

> Artificial Super Intelligence

Artificial Super Intelligence is the stage of Artificial Intelligence as soon as the

capability of computer systems may transcend human beings. ASI is now
currently a hypothetical position as depicted in pictures and science fiction
novels, in which makers have obtained on the planet.

I think that machines are not very far from attaining this Stage taking into
considerations our present-day tempo.

Except if you have a direct introduction to gatherings like Deep personality,

you have no clue how quick – it is developing at a pace near exponential- Elon
Musk cited.

Thus, all these would be exactly the various phases of intellect that a machine
may acquire. Now let us know the forms of all AI, based on your own
functionality.
 Types Of Artificial Intelligence

Whenever Someone asks one to clarify the Different Kinds of Artificial

Intelligence systems, you have to categorize them based on their own
functionalities.

Based on the performance of AI-based Methods, AI can be categorized into the

following varieties:

1. Reactive Machines AI
2. Limited Memory AI
3. Theory Of Mind AI
4. Self-aware AI

Types of artificial intelligence :-

1.PURELY REACTIVE: This one is a most basic system of AI. Its the one
which does not make a stock of memories or any past deals. It reacts directly to
what it sees. It doesn’t have the approach of the wider world.

2.LIMITED MEMORY: After purely reactive in the expansion of AI this

came into exists, which adds up on pieces of past information into future
representations of the world.

3.THEORY OF MIND: This type of AI having the capacity to understand

thoughts and feelings which affect human behaviour. This can deal with
feelings, motives, intentions, expectations in short all types of inner emotions
and can interact socially.

4.SELF- AWARE: Here AI can form representations about themselves. They

are aware of there internal states, that can predict the feelings of others and can
make abstraction and assumptions. They are said to be future generations of
machines with super intelligence.
 4. Branches of Artificial Intelligence

Artificial Intelligence May Be Used to solve Real World problems By

implementing the following processes/ Techniques:

 Machine Learning
 Deep Learning
 Natural Language Processing
 Robotics
 Expert Systems
 Fuzzy Logic

> Machine Learning

Machine Learning will be The science to having machines interpret, analyze

and process data as a way to fix real-world problems.

AI for Lead Generation — Here is another use of artificial intelligence in

today’s time that using AI you can generate automatic sales Leads for your
Business.
 Under Machine Learning you can find 3 categories:

1. Supervised Learning
2. Unsupervised Learning
3. Reinforcement Learning

> Deep Learning

Deep Learning is the process of implementing Neural Networks on High

dimensional information to gain insights and form answers. Deep Learning is a
complex field of Machine Learning that can be used to solve more complex
issues.

Deep Learning is your logic behind the Encounter affirmation Algorithm on

Facebook, self-driving autos, digital assistants like Siri, Alexa and so forth.

> Natural Language Processing

Natural Language Processing (NLP) describes the science of Drawing insights

from pure human language as a way to communicate with computers and also
grow companies.

Twitter uses NLP to filter outside terroristic Speech within their tweets,
Amazon makes use of NLP to comprehend customer critiques and increase user
experience.

> Robotics

Robotics is still a division of Artificial Intelligence which focuses on Various

branches and use of all bots. AI Robots are synthetic agents behaving in a real-
world atmosphere to create consequences by taking accountable actions.

“ Sophia the Humanoid can be Really a Great Illustration of AI in Robotics.”

> Fuzzy-logic

Fuzzy-logic is a calculating strategy based on the principles of “degrees of

truth” instead of the customary modern computer logic i.e. Boolean method.
Fuzzy logic is used from the Health Care areas to solve complex problems that
involve decision making. They are also utilized in automatic gearboxes, vehicle
environment control and so forth.

> Expert Systems

A Specialist system is the AI-based pc system that learns and Reciprocates the
decision-making ability of a person specialist.

Expert systems use if-then logical notations to Solve complex issues. It will not
rely upon traditional procedural Programming. Expert systems are mainly
Utilised in information management, health Centers, loan analysis, virus
detection and thus on.

—IkhlasMohd. Saqib

AI software and tools

Since AI is used in a variety of applications across the spectrum, this best-of list
also covers a wide variety of industries and uses. Some tools perform one
specialized task while others offer more flexibility in how it’s used. This list is
only a small representation of the growing number of tools available today, and
it’s in no specific order. We looked at reviews from multiple sites, features,
customer feedback, and market ranking as our basis, with the aim of helping
you find the right AI software for your needs.

MATLAB
MATLAB from MathWorks is a downloadable programming tool in a desktop
environment. It allows you to create code, algorithms, and applications for
running your data iteratively and automatically. Features include a Live Editor
to create scripts in an executable notebook, tools to organize, clean, and analyze
data sets, a drag-and-drop App Designer, a plot gallery to display graphs in
many ways, and an API to use it with other programming languages such as
Java and Python.

New Relic
New Relic One is a telemetry data platform for software engineers. It collects
all types of operational data such as metrics, events, and logs. It then traces this
data into a single place and allows software engineers to observe, analyze, and
troubleshoot problems across their entire software stack. A built-in applied
intelligence engine automatically detects, diagnoses, and resolves issues, so you
can respond to incidents before your customer does. The platform integrates
with hundreds of tools and open standards to simplify monitoring of telemetry
data.

TensorFlow
TensorFlow is a free, open source software library to build machine learning
(ML) models. It includes tools to build and train ML models such as intuitive
APIs. You can deploy your ML-powered apps whether in the cloud, in servers
on-premise, in browsers, and in other devices. Use cases include training a
neural network to identify specific anatomy during a brain MRI, ranking Tweets
in a timeline view, and predicting the dimension and weight of a shipment.

Best Machine Learning Software

ArcGIS
Esri ArcGIS is a powerful mapping and analytics software. AI software and
location-based analytics allows businesses to gain insights with tools that
visualize, analyze, and contextualize data. Users can collaborate and share
information using maps, apps, dashboards, and reports. It has various use cases,
such as spatial analysis, field operations, mapping, 3D modeling, remote
sensing, and data collection and management.

HOLMES
Wipro HOLMES is an AI and automation platform for business and IT users.
The software supports digital transformation of businesses with solutions that
include natural language processing, algorithmic intelligence, self learning, and
reasoning. Businesses can use the software together with current transaction
systems. Users that derive the most benefit are C-suite executives in charge of
operations, finance, legal department, procurement, HR, and marketing. The
platform also caters to IT end users, DevOps, operations, digital networks,
cloud, cybersecurity, SAP, and Oracle users.

Watson
Watson is IBM’s suite of enterprise-ready AI software applications, services,
and tools. It includes pre-built applications and tools to help you build, run, and
manage AI in your organization. Watson helps free up users from doing
repetitive tasks so they can instead focus on more creative work. They can apply
the software in IT operations, customer service, risk and compliance, and
financial operations. The AI platform can help enrich customer interactions,
anticipate and preempt disruptions, and recommend actions.

Einstein
Einstein from Salesforce is an artificial intelligence software that serves as a
smart CRM assistant within Salesforce software. The AI tool helps Salesforce
users discover insights, predict outcomes, recommend next steps, and automate
workflows. It can identify patterns and trends, anticipate customer actions based
on behaviors, point customers in the right direction, and automate data entry.

Azure AI Platform
Microsoft’s Azure AI Platform enables you to build applications that can
analyze images, comprehend speech, and make predictions based on data. Data
scientists, ML engineers, and app developers are the main users of the software.
AI-powered apps run on a secure network where users can easily scale them
across their infrastructure. Features include machine learning, knowledge
mining, customizable services, and purpose-built bots. Microsoft also offers
Cortana, an AI productivity assistant in Microsoft 365 apps.

Alexa
Amazon Alexa is a cloud-based virtual assistant AI voice service that users can
extend and expand using tools, APIS, and reference solutions. Users can build
Skills, which are the counterpart of apps for Alexa using its Skills Kit. They can
build devices by integrating Alexa into devices and controlling them by voice.
Enterprises can also use Alexa for workplace applications such as hospitality
establishments.

Google Cloud AI
Google offers several AI and machine learning products on its cloud platform. It
has a platform for training, hosting, and managing ML models. AI building
blocks let users add sight, language, conversation, and structured data to
applications that you build. It has conversational AI that can convert speech to
text or text to speech, virtual agents, and DialogFlow for creating chatbots,
voicebots, and IVR bots. Google also has AI products for documentation and
specific industry solutions.

Assignment questions :

1. Name the simulation tools and softwares used in intelligent

manufacturing process.
2. Define Artificial Neural Network? Draw the basic model of
Artificial Neural Network?
3. State and classify the types of Neural Networks? Determine the use
of Artificial Neural Networks?
4. Explain the role of the intelligent systems and their potential
benefits. List the major advantages and disadvantages of artificial
intelligence over natural intelligence.
5. Write the advantages of neural network over conventional
computers.
6. List some practical applications of Artificial Neural Network.
7. List the major advantages and disadvantages of artificial intelligence
over natural intelligence.