CS3101: Artificial Intelligence and Soft

Computing
Introduction

Ashish Kumar
Dept. of CSE
Manipal University Jaipur

1
Introduction
This course introduces artificial intelligence techniques
and soft computing techniques to the students.
The course will teach them about Autonomous Agents,
Problem solving, Search, Heuristic methods, State space
Learning, Game Playing, Knowledge Representation,
Uncertainty, Propositional Logic, Predicate Logic, Logic-
based Agents, Basics of Natural Language Processing,
Neural Networks, Evolutionary Computation.

2
Course Outcomes
1. Describe basics of Artificial intelligence and Intelligent
agent.
2. Define AI problem and apply appropriate AI search
technique to solve the problem.
3. Illustrate knowledge representation using propositional,
first order predicate logic and semantic network and
apply reasoning process to draw conclusions.
4. Apply soft computing and NLP techniques to solve
problems and Improve entrepreneurship skills.
5. Apply evolutionary technique for solving optimization
problems.

3
Syllabus
 Introduction: Fundamental Concepts: Agents, environments, general model,
Problem solving techniques.
 Search Techniques: Uninformed search, heuristic search, adversarial search
and game trees, Solution of constraint satisfaction problems using search.
 Knowledge Representation: Propositional and predicate calculus, semantics
for predicate calculus, inference rules, unification, Resolution, semantic
networks, conceptual graphs/Dependency, structured representation.
 Learning: Inductive learning, decision tree learning.
 Natural language processing: introduction, parsing using context free
grammars, Chomsky hierarchy, case grammar.
 Soft computing: Fuzzy set theory, Fuzzy sets, set-theoretic operations,
membership functions, Union, intersection and complement, fuzzy rules,
reasoning, and interference.
 Neural networks: Perceptron, Back Propagation. Evolutionary techniques:
genetic algorithms, Swarm Algorithm, ant colony optimization

4
Books
Reference Books:
S. Russell, and P. Norvig: Artificial Intelligence: A Modern
Approach. Prentice Hall, 2011.
E. Rich, K. Knight, and S.B. Nair: Artificial Intelligence. 3rd
Ed., Tata McGraw Hill, 2009.
G. F. Luger: Artificial Intelligence-Structures and Strategies
for Complex Problem Solving, (6e), Addison-Wesley Pearson
Education, 2012.

5
Prerequisites
Comfortable programming in language such as C (or C++)
or Java 0r Python.
Some knowledge of algorithmic concepts such as running
times of algorithms; having some rough idea of what NP-
hard means.
Some familiarity with probability (we will go over this from
the beginning, but we will cover the basics only briefly).
Not scared of mathematics, able to do simple
mathematical proofs.

6
Course Evaluation
Component Maximum Marks
Sessional I (MTE-1) 30
Problem Statement 5
CWS Abstract, Literature Review 5
(Paper Methodology 5
Writing) Implementation 5
Comparison/Result & Submission 10
End Semester Exam 40

7
Course Conduct
Maintain 75% attendance.
If you are in class then only attendance will be yours.
Submit your assignment & quizzes within given time
period.
It has your 30 marks.

8
Course Faculty Details

Ashish Kumar
Associate Professor,
Dept. of CSE,
Manipal University Jaipur

Contact No. - +91-8233244583

Email – kumar.ashish@jaipur.manipal.edu

9
What is Artificial Intelligence?
Artificial Intelligence is concerned with the design of
intelligence in an artificial device.
The term was coined by John McCarthy in 1955.
A serious science.
General-purpose AI like the robots of science fiction is
incredibly hard
 Human brain appears to have lots of special and general
functions, integrated in some amazing way that we really do
not understand at all (yet)
Special-purpose AI is more achievable (nontrivial)
 E.g.,chess/poker playing programs, logistics planning,
automated translation, voice recognition, web search, data
mining, medical diagnosis, keeping a car on the road.
10
Some Definitions

11
Some Definitions

12
 if our system can be
more rational than
Definitions of AI humans in some
cases, why not?
focus on action avoids
philosophical issues
such as “is the system
conscious” etc.

• We will follow “act rationally” approach

• Distinction may not be that important
• acting rationally/like a human presumably requires
(some sort of) thinking rationally/like a human,
• humans much more rational anyway in complex
domains
13
Fundamental Concepts
There are two ideas in all the definition.
 Intelligence
 Artificial Device
 What is Intelligence?
 Ability to take decisions(Classify/choose best option/etc)
 Use of knowledge to respond in a new situation
 Inductive inference(ability to generalize)
 Ability to learn(Acquire knowledge)
 Ability to express ourselves (language)
 What is Artificial Device?
 Any object or machine that has been created or designed by humans to
perform specific functions or tasks.
 Such as electronics gadgets, robots, medical equipment, communication
devices, computers.
 Also known “Agent”.

14
Timeline of AI

15
AI vs M.L. vs D.L.

16
Can we judge the Intelligence of Machine?
 The Turing Test, proposed by British mathematician and computer
scientist Alan Turing in 1950
 It is a test of a machine's ability to exhibit intelligent behavior
indistinguishable from that of a human.
 In the Turing Test, a human evaluator interacts with a machine and a
human through a computer interface without knowing which is which.
 If the evaluator cannot reliably tell which responses are from the
machine and which are from the human, the machine is said to have
passed the test and demonstrated artificial intelligence (AI).
 It is definitive measure of intelligence.
 It does not consider other aspects of intelligence,
such as creativity, problem-solving ability,
emotional intelligence, or the ability to learn and
adapt over time.
 http://www.jabberwacky.com/
 https://www.cleverbot.com/ 17
Lessons from AI Research
 Clearly-defined tasks that we think require intelligence and education
from humans tend to be doable for AI techniques.
 Playing chess, drawing logical inferences from clearly-stated facts,
performing probability calculations in well-defined environments, …
 Although, scalability can be a significant issue.
 Complex, messy, ambiguous tasks that come natural to humans (in
some cases other animals) are much harder.
 Recognizing your grandmother in a crowd, drawing the right conclusion
from an ungrammatical or ambiguous sentence, driving around the city, …
 Humans better at coming up with reasonably good solutions in
complex environments
 Humans better at adapting/self-evaluation/creativity (“My usual
strategy for chess is getting me into trouble against this person… Why?
What else can I do?”)

18
Modern AI
More rigorous, scientific, formal/mathematical
Fewer grandiose promises
Divided into many subareas interested in particular aspects
More directly connected to “neighboring” disciplines
 Theoretical computer science, statistics, economics, operations
research, biology, psychology/neuroscience, …
 Often leads to question “Is this really AI”?
Some senior AI researchers are calling for re-integration of
all these topics, return to more grandiose goals of AI
 Somewhat risky proposition for graduate students and junior
faculty…

19
Approaches to AI
 Strong AI aims to build machines that can truly reason and solve problems.
These machines should be self aware, and their overall intellectual ability
needs to be indistinguishable from that of a human being. Excessive
optimism in the 1950s and 1960s concerning strong AI has given way to an
appreciation of the extreme difficulty of the problem. Strong AI maintains
that suitably programmed machines are capable of cognitive mental states.
 Weak AI deals with the creation of some form of computer-based artificial
intelligence that cannot truly reason and solve problems but can act as if it
were intelligent. Weak AI holds that suitably programmed machines can
simulate human cognition.
 Applied AI aims to produce commercially viable "smart" systems such as,
for example, a security system that is able to recognize the faces of people
who are permitted to enter a particular building. Applied AI has already
enjoyed considerable success.
 Cognitive AI computers are used to test theories about how the human
mind works--for example, theories about how we recognize faces and other
objects, or about how we solve abstract problems.
20
Application of AI
 Natural Language Processing (NLP): AI is used in NLP applications
to understand and process human language. Examples include virtual
assistants like Siri and Alexa, language translation services, sentiment
analysis in social media, and chatbots for customer support.
 Image and Speech Recognition: AI is utilized for image recognition
tasks, such as identifying objects, people, and scenes in images and
videos. Speech recognition is applied in virtual assistants, voice-
controlled systems, and speech-to-text conversion.
 Autonomous Vehicles: AI plays a crucial role in self-driving cars and
autonomous drones. It enables these vehicles to perceive their
surroundings, make real-time decisions, and navigate safely without
human intervention.
 Healthcare: AI is employed in medical imaging to assist in diagnosing
diseases and detecting anomalies in X-rays, MRIs, and CT scans. It is
also used for drug discovery, personalized medicine, and virtual health
assistants.
21
Application of AI
 Finance: AI is used in the financial industry for fraud detection,
algorithmic trading, credit risk assessment, and customer service
automation.
 Gaming: AI techniques are employed to create intelligent virtual
opponents in video games, enabling more challenging and realistic
gameplay.
 Recommender Systems: AI-based recommendation engines are used
in e-commerce, streaming platforms, and content websites to provide
personalized suggestions to users based on their preferences and
behavior.
 Robotics: AI is integrated into robots for tasks like industrial
automation, logistics, exploration in hazardous environments, and
assistance for the elderly and people with disabilities.
 Virtual Reality and Augmented Reality: AI enhances immersive
experiences in VR and AR applications by providing intelligent
interactions and content personalization.
22
Application of AI
 Climate Modeling: AI is utilized in climate science for predictive
modeling, weather forecasting, and analyzing climate data to better
understand and address climate change.
 Cybersecurity: AI is employed in cybersecurity to detect and prevent
cyber threats, identify patterns of malicious behavior, and enhance
network security.
 Education: AI is used in educational applications to personalize
learning experiences, provide intelligent tutoring, and assess student
performance more effectively.

23
Current Trends in AI
 Advancements in Deep Learning: Deep learning, a subset of AI,
continued to drive significant progress in various domains, including
computer vision, natural language processing, and speech recognition.
Researchers were exploring larger and more sophisticated deep neural
network architectures to improve performance on complex tasks.
 Explainable AI (XAI): As AI models became more complex, there was
a growing need for interpretability and transparency. Explainable AI
aims to provide insights into how AI models arrive at their decisions,
making AI more trustworthy and facilitating regulatory compliance.
 AI in Edge Computing: With the proliferation of Internet of Things
(IoT) devices, AI was increasingly being deployed at the edge of
networks. This approach reduced latency, improved privacy, and made
real-time decision-making possible without relying heavily on cloud
infrastructure.

24
Current Trends in AI
 Reinforcement Learning (RL) Advancements: RL, a type of
machine learning where agents learn to take actions to maximize
rewards, saw significant developments. It found applications in
robotics, autonomous vehicles, and gaming, among others.
 AI Ethics and Responsible AI: The responsible and ethical use of AI
gained more attention. Researchers and organizations were focusing
on addressing bias, fairness, and privacy concerns in AI systems.
 AI in Healthcare: AI applications in healthcare continued to grow,
with AI assisting in medical diagnosis, drug discovery, personalized
treatment plans, and healthcare data analysis.
 AI in Natural Language Processing (NLP): NLP technologies
witnessed rapid advancements, enabling more sophisticated language
understanding and generation, as well as improving conversational AI
and language translation systems.

25
Current Trends in AI
 AI in Autonomous Vehicles: The development of self-driving cars
and autonomous vehicles gained momentum, with AI algorithms
playing a crucial role in perception, decision-making, and navigation.
 AI and Creativity: AI's role in creative tasks, such as art generation,
music composition, and content creation, garnered interest and
showcased the potential of AI to augment human creativity.
 AI for Climate and Sustainability: AI was increasingly being applied
to address climate change and sustainability challenges, including
optimizing energy consumption, environmental monitoring, and
climate modeling.
 Federated Learning: Federated learning, a privacy-preserving
approach to train machine learning models across multiple devices or
servers without sharing raw data, gained popularity as a way to protect
user privacy.

26
Current Trends in AI
 Democratization: low-code, no-code AI: AI is becoming more
accessible to a wider range of people through the development of low-
code and no-code AI platforms. These platforms allow people to build
and deploy AI models without having to be experts in machine
learning. This is making AI more accessible to businesses and
individuals who may not have the resources to develop their own AI
models.
 Digital twinning: Digital twinning is a process that creates a virtual
replica of a physical object or system. This can be used to test and
simulate new products or systems, to improve the performance of
existing systems, or to predict failures.

27
Thank You

28