Course Objectives

CS210: Artificial Intelligence

• A brief idea to the philosphy and breadth of ideas in AI
• Basic ideas and techniques underlying the design of
intelligent computer systems / agents.
• Learn the representation and use of knowledge in
Dr. Chandra Prakash inference-based problem solving approaches
Assistant Professor • Learn to apply probability theory to describe and model
Department of Computer Science and Engineering agents operating in uncertain environments
• Learn statistical and decision-theoretic modeling
paradigm.
• Presentation practice

(Slides adapted from StuartJ. Russell, B Ravindran, Mausam, Prof. Pallab Dasgupta, Prof. Partha Pratim Chakrabarti, Saikishor
1 5
Jangiti, Shivaram Kalyanakrishnan

Syllabus Reading Material

• Module 1: Introduction to AI For students who want to read more we recommend

• Module 2: Automated Problem Solving

• Module 3: Logic and Deduction

• Module 4: Planning in AI

• Module 5: Reasoning Under Uncertainty

• Module 6: Machine Learning

• Module 7: AI Applications

• On line relevent material will be shared with you

6 7
 Course website Course Policies
• Website : • Lecture notes, programming assignments, and other useful information will be posted
on the course web page.
– tentative schedule update
• You should check the web page regularly.
– lecture slides and notes
• The lecture notes will be in power point.
– course policies, etc. • Discussion of the programming assignments is allowed and encouraged. However, each
– https://cprakash86.wordpress.com/csb-210-ai-2024/ team is expected to do its own work.
– Discussion (lecture related, doubts) • Assignments which are similar will receive a zero.
– [email protected] • Regular attendance is highly recommended. If you miss a class, you are responsible for
all material covered or assigned in class. Late programming assignments will be
– Teaching Assistant [TA]:
penalized 10% of the points assigned per day (weekends count as one day).
•

8 9

Course Components About ‘the’ Course

• Continuous Assessments [ 20] :
– An assignment based course
– Attendance
– Surprise Quizzes – More emphasis on developing an
– projects, groups of 4-5 solution for a real time problems
• Academic integrity! – Peer learning through presentation
• Python – Project:
• Give you hands-on experience with the algorithms
• Using Programming in Prolog and
• Mid Term [30] Python
• End Term [50] – Pre-requisites
• Data Structures
• LAB: • Probablity
• Programming assignments 10 11
 Form Analysis Your Expectation
• Hope this course teaches a lot about the field of AI both theoretically and practically, so that by
the end of this course we are at least able to make sone mini projects on our own
• I should be able to develop ai solutions
• New learnings with innovations
• To have a good knowledge at ML / Able to build own machine learning model.
• Notes+worksheets+real life applications examples
• To provide course material or reference books for a particular unit after it has been taught
• Would love to have teacher student interactions, access to ask doubts, and receive notes, ppts or pdfs
• Team quizzes and assignments where groups of students work together on problems or even
assignments building coordination and understanding amongst each other.
• Study Reinforcement Learning, a key aspect of AI involving decision-making and optimization.
• I do not want to attend boring lecture mean by completion of syllabus but it should be more
about concepts.

12 13

Project : Technology Readiness Levels (TRLs) How to keep motivated in this course:
• Technology Readiness Levels (TRLs)
are a method for understanding the
technical maturity of a technology
during its acquisition phase. Don't Search for Numbers,
• TRLs allow engineers to have a
consistent datum of reference for
Search for your queries/Answers
understanding technology evolution,
regardless of their technical
background.

“I would rather have questions that can't

be answered than answers that can't be
Source: https://www.twi-global.com/technical-knowledge/faqs/technology-readiness-levels
14
questioned.” ― Richard P. Feynman 15
 Module 1: Introduction to the Computers Human Sapiens
• PART 1.1: What is Artificial Intelligence • Latin-
• PART 1.2: History of AI – wise man

• PART 1.3: Possible Approaches in AI • Trying to understand How we

• PART 1.4 : Application Domains and brief Overview of Modern AI – think
– act
• PART 1.5: Areas Contributing to AI

• PART 1.6 : Core Capabilities covered in this course • Are humans the only intelligent species?
• Do we include all living beings as intelligent?
– There do not exist standard and mathematically precise definitions of
16 intelligence. 17

Artifical Intelligence (AI) Imagination / Reality

• ?????

• Artificial = Machine
• But what is intelligence?
• AI is the an attempt of reproduction of human reasoning and
intelligent behavior by computational methods

Modern-day videoconferencing[2]

18 From the Mahabharata[1] 19

 Sci-Fi AI? Today's Robot

20 21

Lets Start Learning / use Brain?

§ Brains (human minds) are very good
at making rational decisions, but not
2403343781289312
perfect
+
+
2843033712837981
2362142787897881
§ Brains aren’t as modular as software,
+ 3256541312323213 so hard to reverse engineer!
+ 9864479802118978
+ 8976677987987897 § “Brains are to intelligence as wings
+ 8981257890087988 are to flight”
=?
§ Lessons learned from the brain:
memory and simulation are key to
decision making
We are naturally good at certain things.
○ Pattern recognition. ○ Languages and speech. ○ Reasoning. ○ Planning. ○ Learning. ○ Exploring. 23
 Human Cognition Abilities What is Intelligence ?
• Intelligence (Oxford dictionary ):
• Ability to
• Learn
• Understand and
• Think.

24 25

What is involved in INTELLIGENCE Artificial + Intelligence

• Ability to interact with the real world
– to perceive, understand, and act
• Artificial :
– e.g., speech recognition and understanding and synthesis o fake, not real , man made
– e.g., image understanding o aeroplane= artifical flying
– e.g., ability to take actions, have an effect o ships= artifical swiming
• Reasoning and Planning
– modeling the external world, given input
• Intelligence:
– solving new problems, planning, and making decisions – “the capacity to learn and solve problems”
– ability to deal with unexpected problems, uncertainties – in particular,
• Learning and Adaptation • the ability to solve novel problems
– we are continuously learning and adapting • the ability to act rationally
– our internal models are always being “updated” • the ability to act like humans
• e.g., a baby learning to categorize and recognize animals
26 27
 Computer Systems What is “Artificial Intelligence”?
• Common in almost all aspects of our daily lives. – Alan Turing [ 1950]
– Hard to imagine a world without them. – I proposed to consider the question:
can machine think ?
• [A. M. Turing (1950) Computing Machinery and
Intelligence. Mind 49: 433-460.]

• Turing Test & Total turing test [1950]

• Operational test to determine an entity is intelligent /
not
– The computer is interrogated by a human via a teletype.
– It passes if the human cannot tell if there is a computer or
human at the other end
28 29

1955-56 Birth of the word AI Some Early successes of Dartmouth

• Term coined by, John McCarthy (1955) • Many key projects were initiated after dartmouth
– AI – “the science and engineering of making intelligent machines” summer project.
a) Shakey robot [1966 - 1972]
• Dartmouth Summer Research Project on Artificial Intelligence – Combined research in robotics, computer vision and
(1956) natural language processing
– First mobile robot to perceive environment
– Could reason about its surroundings and actions!
• Introduced A* algorithm to find paths
• Hough Transform for image analysis
• Used Lisp for programming
• visibility graph used for finding shortest paths
in the presence of obstacles.

30 31
Source : http://www.ai.sri.com/shakey/
 Some Early successes of Dartmouth ARTIFICIAL INTELLIGENCE
b) Dendral There are no clear agreement on the definition of AI
– attempted to encode the domain expertise in molecular biology as an expert • It is the science and engineering of making intelligent machines,
system especially intelligent computer programs.
– determining 3D structures of complex chemical compounds
• It is related to the similar task of using computers to understand human
– Led to the creation of expert systems for various other domain, including
intelligence, but AI does not have to confine itself to methods that are
medical.
biologically observable.
• A milestone in the history of AI !!!
• AI is the study of how to make computers just like humans. That means
how to make computers to do things that people do better.

32 33

Other possible AI definitions What is AI?

• AI is a collection of hard problems which can be solved by humans and other The science of making machines that:
living things, but for which we don’t have good algorithms for solving.
– e. g., understanding spoken natural language, medical diagnosis, circuit
design, learning, self-adaptation, reasoning, chess playing, proving math
theories, etc. Think like people Think rationally
• AI is a process of making a machine or a program that
AI tends to work
– Learn and understand like human mostly in this area
– Acts like human (Turing test)
– Thinks like human (human-like patterns of thinking steps)
Act like people Act rationally
– Acts or thinks rationally (logically, correctly)

34 35
 Rational Decisions Maximize Your Expected Utility
We’ll use the term rational in a very specific, technical way:
§ Rational: maximally achieving pre-defined goals
§ Rationality only concerns what decisions are made (not the thought process behind them)
§ Goals are expressed in terms of the utility of outcomes
§ Being rational means maximizing your expected utility

36 37

Designing Rational Agents Cont…

• An agent is an entity that perceives and acts. AI is the study and design of intelligent agents
• A rational agent selects actions that maximize its
(expected) utility.
• An agent is a function from percept histories to
where,
actions:
[f: P* → A]
an intelligent agent is a system that interact with its environment and
• Characteristics of the percepts, environment, and
action space dictate techniques for selecting rational
takes actions that maximize its chances of success.
actions
Environment

Sensors
• This course is about: Percepts
Agent

– General AI techniques for a variety of problem

types ?
– Learning to recognize when and how a new
problem can be solved with an existing technique Actuators
Actions
38 39
 Problems In AI Humanoid robot
Easy Problems in AI
• It’s been easier to mechanize many of the high level cognitive tasks we usually
associate with “intelligence” in people
– e. g., symbolic integration, proving theorems, playing chess, some aspect of
medical diagnosis, Engineering tasks, Financial, etc.
Hard Problems in AI
• can be solved by humans and other living things, but for which we don’t have good
algorithms for solving
• It’s been very hard to mechanize tasks that animals can do easily
– walking around without running into things (ASIMO)
– Perceptual task :
• interpreting complex sensory information (visual, aural, …)
– working as a team (ants, bees)
40 41
• Algorithmic view such as NP hard , Search , Game playing, planning

AI pioneers Foundations of AI
• Alan Turing(1912-1954) • Philosophy: Logic, methods of reasoning, mind as physical system,
foundations of learning, language, rationality.
– Father of computer science
• Mathematics: Formal representation and proof, algorithms,
– Turing test for AI computation, (un)decidability, (in)tractability

• Marvin Minsky (MIT) – 1956 • Probability/Statistics : modeling uncertainty, learning from data
• Economics : utility, decision theory, rational economic agents
– Built first Neural network computer SNARC • Neuroscience : neurons as information processing units.
• John McCarthy ( Stanford University ) • Psychology / Cognitive Science : how do people behave, perceive,
process cognitive information, represent knowledge.
– Developed LISP, AI programming language
• Computer : building fast computers engineering, GPU,TPU, Wafer
• 2018 Turning Award for Deep Learning scale engine (WSE) , Quantum computing
• Control theory: design systems that maximize an objective function
– Jefrrey Hinton, Yoshua Bengio, Yann LeCun over time
• Linguistics : knowledge representation, grammars

42 43
Confusion [ AI, ML, DL ] ???

2020

bl e ge
um gui gua
m h t i n l an
an s ha
fro indis uces
ost rod
alm PT3 P
tas AI's G
ks
en
Op
44 45

A brief history of AI Programming languages for AI

• The programs for AI problems can be written with on procedural languages like
PASCAL or declaration languages like PROLOG.
• Generally relational languages like PROLOG or LISP are preferred for symbolic
reasoning in AI.
• If the program requires much arithmetic computation (say for the purpose of
uncertainty management), then procedural languages would be preferred.
– Python. • C++
– Prolog. • JavaScript.
– LISP. • Java.
– R. • Haskell.

46 47
 Dimensions / Possible Approaches in AI Thinking Humanly
• Cognitive Modelling Approach
Like humans Well/ Rational – very hard to understand how humans think Like
– Requires scientific theories of internal activities of the human brain humans Well
– How do we capture human thinking to implement ?
• Cognitive Science, Neuroscience, Psychology Rational
Rational Think GPS
GPS agents
Think agents • Computational model should reflect “how” results were obtained.

AI tends to work • System : “General Problem Solver (GPS)” (Newell and Simon, 1961) Eliza Heuristic
Act systems
mostly in this area – Designed to work as a universal problem solver
– Problems represented by horn clauses
– First AI Machine which has KB + Inference separation
Heuristic – Goal produce a sequence of steps of the reasoning process that was similar
Act Eliza systems to the steps followed by a person in solving the same task.

• Growth of Cognitive science and AI supports each other

• A machine that thinks like human while solving a problem correctly

48 49

Thinking well / rationally Acting Well / Rationally Like

• Law of thoughts Approch • Acting Rationally- The Rational Agent Approach humans Well
– Greek Philosopher Aristotle, Third century BC. – right thinking
Like – rational behavior = doing the right thing
humans Well – An agent is an entity that perceives and acts Rational
– Belief that “logic” governs the human thought process Think GPS
– This course is about designing rational agents agents
• Develop formal models of knowledge representation, reasoning, Rational
Think GPS
learning, memory, problem solving, that can be result in agents • Abstractly, an agent is a function from percept histories to actions:
algorithms. [f: P* → A] Eliza Heuristic
Act systems
– There is often an emphasis on a systems that are provably correct, and guarantee
finding an optimal solution. Heuristic
Act Eliza
systems
• Design best program for given machine resources
• Thinking Rationally – For a given set of inputs, generate an appropriate output that is not necessarily correct but gets the job
done.
– Field of Logics gave rise to codifying rational thinking
– When elements are ‘things’, we reason about things
• A heuristic (heuristic rule/method) is a rule of thumb, strategy, trick, simplification, or any
• Hurdles to the idea : other kind of device which drastically limits search for solutions in large problem spaces.
– Not everything can be logically coded : eg Hot Fire , reflix – Heuristics do not guarantee optimal solutions; in fact, they do not guarantee any solution at all
– No provably correct action at a moment – All that can be said for a useful heuristic is that it offers solutions which are good enough most of the
time.
– Exhaustive computational resources
50 51
 Act like humans Eliza
• Behaviorist approach. Like • Joseph Weizenbaum, 1964
humans Well • ELIZA: A program that simulated a psychotherapist
• Not interested in how you get results, interacting with a patient and successfully passed the
just the similarity to what human results Think GPS Rational Turing Test.
agents • Coded at MIT during 1964-1966 by Joel Weizenbaum.
are.
• Natural Language Processing Computer Program
• Exemplified by the Turing Test (Alan • First Chatbot!
Heuristic
Turing, 1950). Act Eliza
systems • First script was DOCTOR as Psychotherapist
– The script was a simple collection of syntactic
patterns not unlike regular expressions
– Each pattern had an associated reply which might
include bits of the input (after simple transformations
(my ® your)

Source : https://news.mit.edu/2008/obit-weizenbaum-0310

52 53

Acting Humanly: The Turing Test The Loebner Contest

• Alan Turing (1912-1954)
• A modern version of the Turing Test, held annually, with a $100,000 cash
prize.
– [ http://www.loebner.net/Prizef/loebner-prize.html ]
• Restricted topic (removed in 1995) and limited time.
• Participants include a set of humans and a set of computers and a set of
judges.
• Scoring
• The interrogator is limited to using the responses to written questions in order to make the determination.
– Rank from least human to most human.
• Skills necessary to pass these tests
– Highest median rank wins $25000.
– NLP, Knowledge Representation, Automated Reasoning, ML
+ – If better than a human, win $100,000. (Nobody yet…)
– Computer Vision & Robotics(for total turing test)
54 56
 Chinese Room argument (1980) Cont…
• Devised by John Searle
• Argument against the possibility of • Some problems used to be thought of as AI but are now considered
true artificial intelligence. not
• The argument centers on a thought – e. g., compiling Fortran (suited to numeric computation and scientific
experiment in which someone who computing) in 1955,
knows only English sits alone in a
room following English – symbolic mathematics (manipulate mathematical equations ) in 1965
instructions for manipulating – proving math theories
strings of Chinese characters, such
that to those outside the room it
appears as if someone in the room
understands Chinese.

57 58

Some Revised definitions of artificial intelligence FOCUS AREAS FOR AI INTERVENTION in India?
• Preventive and affordable Healthcare
• Education and Skilling
• Agriculture and Rural Development
• Smart Mobility and Intelligent
Transportation Systems
• Retail
• Manufacturing
• Energy management
• Smart Cities and Infrastructure

59 60
Source: Niti Aayog Discussion Paper on AI, June 2018
 Intelligent Systems in Your Everyday Life Maps and Self-driving cars
• Natural Language
– Speech technologies (e.g. Siri) – Automatic speech recognition (ASR)
• Text-to-speech synthesis (TTS)
• Dialog systems
– Language processing technologies
• Question answering : 2011 IBM's Watson
• Machine translation

• Web search
• Text classification, spam filtering, etc…

61 62

Smart Systems Robot

63 64
 Business Intelligence Bionomics
• Fraud Detection
• News generation

65 66

Sophia : Social humanoid robot How is the AI of this Era is Different?

• Activated in 2015, Sophia is known for her
human-like appearance and behavior.
• Developed by Hong Kong-based company Hanson
Robotics.

• In October 2017, the robot became a Saudi

Arabian citizen, the first robot to receive World's first robot citizen's India
citizenship of any country. debut at IIT Bombay

• Sophia uses artificial intelligence, facial and

voice recognition, and visual data processing, is
able to imitate human facial expressions and
gestures, and can make conversation on
predefined topics.
67
 Other Modern AI Factors Vision and Deep Learning
• Machine Learning
– The main driver of recent successes in AI
– Move from ”code” to ”data”

• Continued expansion of open source AI, especially in Python, aiding

machine learning and big data ecosystems.
• Leading deep learning libraries open-sourced, allowing further adoption
by industry.
• Open sourcing of large datasets of millions of labeled images, text
datasets such as Wikipedia has also driven breakthroughs.
69

Applications Computer Vision

Face Recognition Image quality enhancement Gesture Recognition

Beautification Classification

Few Popular Applications: Precision Agriculture, Learner Profiling, Video Captioning, Exploring Patterns from Satellite images, Image detection Object detection (Self driving car) 72
in Healthcare, Identifying specific markers in Genomes, Creating Art and Music, Recommendations, behavior prediction,
 Sequence Modeling Algorithms Game Agents

Pong Enduro Beamrider

73

Game Agents AlphaGo

• Classic Moment: May, '97: Deep Blue vs. Kasparov
– First match won against world champion
• AlphaGo is the first computer program to defeat a professional
– “Intelligent creative” play human Go player
– 200 million board positions per second • At the opening move in Chess there are 20 possible moves. In Go
– Humans understood 99.9 of Deep Blue's moves the first player has 361 possible moves
– Can do about the same now with a PC cluster
• Policy network -selects the next move to play.

• 1996: Kasparov Beats Deep Blue

“I could feel --- I could smell --- a new kind of intelligence across the table.” • In late 2017, AlphaZero, a single system that taught itself from
scratch how to master the games of chess, shogi, and Go, beating
• 1997: Deep Blue Beats Kasparov
a world-champion program in each case.
“Deep Blue hasn't proven anything.”
• AlphaGo must restrict Breath and Depth of search among all
• AlphaGo (2016) Beat Top Human at Go board configurations with heuristics information supplied by
training and winning policy for max reward.

76
Text from Bart Selman, image from IBM’s Deep Blue pages
 Motion Analysis Robotics
• Robotics
– Part mech. eng.
– Part AI
– Reality much
harder than simulations!

• Technologies
Move: Walk Around challenge, one of the official challenges in the NeurIPS 2019
– Vehicles
• 132 miles DARPA Grand
challenge
– Rescue
– Help in the home
– Lots of automation…

• In this class:
– We ignore mechanical aspects
– Methods for control

Images from UC Berkeley, Boston Dynamics, RoboCup, Google

Robots: Human-AI Interaction What can’t AI systems do yet?

80
 What Can AI Do? What can’t AI systems do yet?
Quiz: Which of the following can be done at present?
• Common Sense Problem-Solving and Reasoning
• Play a decent game of KBC?
• Emotional Intelligence
• Win against any human at chess?
• Win against the best humans at Go? • Creativity and Innovation
• Play a decent game of tennis? • Generalization in Unseen Scenarios
• Grab a particular cup and put it on a shelf?
• Unload any dishwasher in any home? • Ethical Decision-Making
• Drive safely along the highway? • Understanding Natural Language at Human Level
• Drive safely along Narela market ?
• Buy a week's worth of groceries on the web? • Dynamic Learning and Adaptation
• Buy a week's worth of groceries at D-Mart? • Cross-Domain Learning
• Discover and prove a new mathematical theorem?
• Perform a surgical operation?
• Unload a know dishwasher in collaboration with a person?
• Translate spoken Chinese into spoken English in real time?
• Write an intentionally funny story?
81 82

RoboCup 2050 challenge, Stages of Industrial Revolution

Source : https://www.youtube.com/watch?v=_PC-V5GJP6Q
Source : https://www.textiletoday.com.bd/industry-4-0-a-typical-discussion-and-recommendation-for-bangladesh-spinning-industry/

• RoboCup and Its Role in the History and Future of AI

– Source : [https://ai.sony/blog/blog-006/ ] 83 84
 AI Good or Bad AI Good or Bad ??
• Possible Risks :
– Lethal Autonomous weapons
– Surveillance and persuasion
– Biased decision making
– Impact on employment
– Cybersecurity
• Robots waking up and deciding to revolt against humanity
– Is it possible ???
– Never happen if we give them the right objective, because there is no such thing as a robot
deciding to change its objective
– how to give them the right objective to begin with
• Colobration between AI and Human
86

History of AI in one shot AI Problem Areas /Tasks

• 1st Generation of AI : • 2nd Generation : Expert • 3rd Generation of AI :
Fomal cognitive Tasks Tasks Perceptual Tasks
– Game – Knowledge Represenation – Perception
• Tic-Tac-Toe – Enginnering • Vision
• Chess • Design • Speech
• Checkers • fault finding – Natual Language
• Go • Manufacturing planning • Understanding
– Mathematics – Medical • Generation
• Logic • Diagnosis • Translation
• Geometry • Medical Image Analysis – Robot Control
• Calculus
– Finanical
• Proving properties of
• Stock market prdeictions
programs
87 89
 Course Plan Bridging the gap
• A brief idea to the philosphy of AI
• A brief ides to breadth of ideas in AI
• General computer scientist
– general tools to aid in atacling a new problem
• Serious AI enthusisast
– a primer from which to launch advanced study

• Theory vs. Modeling vs. Algorithm vs. Applications

– Lecture focused towards modeling
– Assignment tilted towards applications
– Few theorems
90 91

Core Capabilities: Covered in this course

• The ability to solve problems
– Searching Algo, Constraint satisfaction, Optimization
• The ability to plan
we can not cover everthing – Abstraction
• The ability to deduce
– Logic, Reasoning algorithms
• The ability to learn
– Models, Data, Learning algorithms
• The ability to handle uncertainty
– Probabilistic and Neural
92
• The ability to interface with the real world 93
 Constraints and Optimization AI Planning
• Path Finding • Elements of a Planning Problem
– I wish to find a shortest path – A set of states (worlds) described in terms of
– I wish to find a path with minimum congestion predicates
– I wish to find a path with combination of – A set of actions which transforms some parts of
transportation options (metro, bus, taxi) one world to take us to another world
– I wish to find a path which goes past a medicine – An initial world
shop – A goal in terms of the predicates that must hold
– I wish to find a path which minimizes energy in the final world
consumption from my battery in a e-vehicle • Planning is widely used in robotics and automated
• When the size and complexity becomes too big we use control
“heuristic functions” to cut out unnecessary parts. • Modern AI explores techniques that combine
• In the lack of domain knowledge, we can statistically planning with machine learning
learn the best way (reinforcement learning) by – Autonomous driving is one of many areas
exploration. where such combinations are highly relevant
• Modern AI aims to combine learning from data with
structured use of domain knowledge. 94 95

Logical Reasoning Logical Reasoning with Other Fun Things

• Facts • Who is the maternal great uncle of Freya?
• Automated ways to use what is known to reason
about something which is not explicitly known. • Rules
• Automated Reasoning:
– grandfather, grandmother,
– Deduction
– maternalgrandfather, maternalgrandmother ,
• Rule: All the marbles in this bag are blue
• Case: These marbles are from this bag • Query :
• Inference: These marbles are blue – maternalgranduncle
– Abduction • Five color theorem: All maps can be colored with
• Rule: All the marbles in this bag are blue five colors, where neighboring countries get
• Observation: These marbles are blue different colors [Proved in 1800s]
• Case: These marbles are from this bag • Four color theorem: All maps can be colored
with four colors, where neighboring countries get
– Induction We need that a social media platform to suggests Freya to post a picture of Fergus
different colors [Proved in 1976 with help of
• Case: These marbles are from this bag computers] on the Maternal-Great-Uncle day
• Observation: These marbles are blue • Applications of Logical Reasoning
• Rule: All the marbles in this bag are blue – Automated Theorem Proving
– Rule-based Systems 96 97
– Complexity Analysis
 Logical Reasoning with Other Fun Things Reasoning under Uncertainty
• grandfather, grandmother, maternalgrandfather, • Who is the maternal great uncle of Freya?
maternalgrandmother , maternalgranduncle

• father( x, z ), father( z, y ) ⇒ grandfather( x, y )

• father( x, z ), mother( z, y ) ⇒ maternalgrandfather ( x, y )

• mother( x, z ), father( z, y )
⇒ grandmother( x, y )

• mother( x, z ), mother( z, y ) • maternalgrandmother( Freya, Charlotte ),

⇒ maternalgrandmother( x, y ) mother( Charlotte, Lindsey ), son( Lindsey,
Fergus ) ⇒ maternalgreatuncle( Freya, Fergus )
• maternalgrandmother( x, z ), mother( z, p ), son( p,y )
⇒ maternalgreatuncle( x, y )
98 99

Machine Learning Course Plan

• A brief idea to the philosphy of AI
• A brief ides to breadth of ideas in AI
• General computer scientist
– general tools to aid in atacling a new problem
• Serious AI enthusisast
– a primer from which to launch advanced study
• Theory vs. Modeling vs. Algorithm vs. Applications
– Lecture focused towards modeling
– Assignment tilted towards applications
– Few theorems
102
 Takeaways Conclusion
• Defination : Intelligence and Artifical Intellignce (AI )
• Artificial Intelligence is a very broad and flexible concept.
– AI Paradox: Once we understand how X works, X is no longer AI!
• Learning AI at a practical level is about methods associated with AI • History of AI : Turning Test, Chines Room Argument
goals.
• Approches in AI
• Arguably, everything a computer does is AI, at a conceptual level. • Application and domains and overview of Modern AI
Traditionally,AI goals have centered around definitions of • Core capabilities of AI
intelligence that go beyond “useful.”
• Robots waking up and deciding to revolt against humanity
• (Arguably, all or almost all statements are arguable.) • Colobration between AI and Human
• The subject of AI deals more with symbolic reasoning that conventional number crusting problems.
• Common areas covered under AI
– Knowledge representation, learning, speech and uncertainty management of data and knowledge.
103 • Python and PROLOG are the used for programming AI problems.

Module 2: Automated Problem Solving Home work 1

• PART 2.1: Intelligent Agent & Environment Due date : 20-Jan
1. Read the Niti Aayog Discussion Paper on AI, June 2018. Find the key areas for AI
• PART 2.2: Problem solving Agent intervention in India. Consider one area and suggest how would you contribute in the
• PART 2.3: Problem Solving Methods selected area (at least 2 pages ) .
• PART 2.4: Search Strategies 2. Read Turing's original paper on AI (Turing, 1950). In the paper, he discusses several
potential objections to his proposed enterprise and his test for intelligence.
• PART 2.5: Adversarial Search a) Discuss the "Heads in the Sand" Objection mention in his article.
• PART 2.6: Constraint Satisfaction Problems b) According to you which objections still carry some importance?
c) Can you think of new objections arising from developments since he wrote the
paper for the next 50 years ?
3. Are reflex actions (such as flinching from a hot fire) rational? Are they intelligent?
Justify your answer.

105 For more questions refer Course webpage 106

 References
• Artificial Intelligence – A Modern Approach, Stuart J Russell, Peter Norvig, Pearson Education
India
• Slides adapted from CS188 Instructor: Anca Dragan, University of California, Berkeley
• Slides adapted from CS60045 ARTIFICIAL INTELLIGENCE

107