Fair Use Notice

The material used in this presentation i.e., pictures/graphs/text, etc.

is solely intended for educational/teaching purpose, offered free of
cost to the students for use under special circumstances of Online
Education due to COVID-19 Lockdown and may include
copyrighted material – the use of which may not have been
specifically authorized by Copyright Owners. It’s application
constitutes Fair Use of any such copyrighted material as provided
in globally accepted law of many countries. The contents of the
presentation are intended only for the attendees of the class being
conducted by the presenter.
Agent Based Intelligent
Systems
Course Learning Outcomes:

Upon successful completion of this course, the student

will be able to:

CLOs Description Taxonomy PLO

level
1 Explain basic principles of Agent based C3 1
Intelligent Systems, related theory and
terminology.
2 To understand and analyze NLP and C4 2
NLP based techniques
3 C5 3
Create agent-based systems for different
computing problems.
What is Intelligence?

 The concept of intelligence has been a widely debated topic

among members of the psychology.
 Intelligence is the capacity of flexible adjustment.
 One could certainly define intelligence by the properties one
exhibits:
 the ability to acquire and apply knowledge and skills.
 the ability to deal with new situations, solve problems, answer
questions or devise plans.
 Intelligence is all about understanding.
 In simple words, intelligence is nothing but thinking skills and the
ability to adapt to and to learn from life’s everyday experiences.
Intelligence can also be observed in terms of:

 Creativity
 Memory
 Reasoning
 Perception
 Problem Solving
 Learning
 Communication
 Adaptation
 Logic
Intelligence can be considered the ability to:
 Learn or understand from experiences
 Make sense out of ambiguous and contradictory messages
 Respond quickly and effectively to a new situation
 Deal with complex situations
 Apply knowledge to manipulate the environment

 Intelligence does not necessarily mean how fast information is

processed, but it is the ability to demonstrate intelligence by
communicating effectively (by any means) and by learning new
concepts (by any means).
Artificial Intelligence:
Definition

 It is the science and engineering of making intelligent

machines, especially intelligent computer programs.

 Artificial intelligence is the study of systems that act in a way

that to any observer would appear to be intelligent.
Artificial Intelligence
Definition
 Artificial intelligence (AI) is wide-ranging branch of computer
science concerned with building smart machines capable of
performing tasks that typically require human intelligence.
 Typically, AI systems demonstrate at least some of the following
behaviors associated with human intelligence: planning, learning,
reasoning, problem solving, knowledge representation, perception,
motion, and manipulation and, to a lesser extent, social intelligence
and creativity.
 AI is an interdisciplinary science with multiple approaches but
advancements in machine learning and deep learning are creating
a paradigm shift in virtually every sector of the tech industry.
Categories of AI
 Weak (narrow) AI embodies a system designed to carry out
one particular job. Weak AI systems include video games,
industrial robots and personal assistants such as Amazon's
Alexa and Apple's Siri. You ask the assistant a question, it
answers it for you.
 Strong AI systems are systems that carry on the tasks
considered to be human-like; they replicate the cognitive
abilities of the human brain. These systems tend to be more
complex as they are programmed to handle situations in which
they may be required to solve problems autonomously (without
human intervention).
What is easy and what is hard for AI?
 It has 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, etc.

 It has been very hard to mechanize tasks that animals can

do easily
• catching prey
• interpreting complex sensory information (visual, aural..)
• modeling the internal states of other animals from their
behavior
• working as a team (ants, bees)
AI – The Ultimate Goal

The ultimate goal of AI is to create systems that:

• Act like humans (Turing test)
• Think like humans (human-like patterns of thinking steps)
• Act and think rationally (logically, correctly)
 Artificial
vs
Natural Intelligence
 Artificial Intelligence:
 AI is more permanent. Natural intelligence is perishable from a commercial
standpoint in that human can change their information. AI is permanent as
long as the computer systems or programs remain unchanged.

 AI offers ease of duplication and distribution. Transferring a body of

knowledge from one person to another usually requires a lengthy process,
yet fully expertise can never be transfer. However, knowledge embodied in
computer systems can be copied or duplicated to another and so on.

 AI can be less expensive that natural intelligence. Some times buying

computer software costs less than having corresponding human power to
carry out same task.

 AI can be documented. Decisions made by a computer can be easily

documented by tracing the activities of a system, while natural intelligence is
difficult to trace out.
Natural Intelligence:
 Natural Intelligence is creative, while AI is uninspired. The ability to
acquire knowledge is inherent in human nature, but with AI customized
knowledge must be built into a carefully constructed system.

 Natural intelligence enables people to benefit from and use sensory

experience directly, while AI mostly works on symbolic inputs.

 Natural intelligence is able to make reasons at all times by wide

context of experience and bring it to bear on individual problems.
While AI systems typically gain their power of knowledge by having a
narrow focus of problem domain.

 Natural Intelligence is powerful but has limitations. Humans are

intellectual but have limited knowledge bases, and information processing
is comparably slow in comparison to computers.
How AI Works:

 Think well

 Act well

 Think like humans

 Act like humans

Think well
 Develop:
 Formal models of knowledge representation
 Reasoning
 Learning
 Memory
 Problem solving that can be rendered in algorithms.

 There is often an emphasis on systems that are provably

correct, and guarantee finding an optimal solution.
Act well

 For a given set of inputs, generate an appropriate output that

is not necessarily correct but gets the job done.

 A heuristic (heuristic rule, heuristic method) is a rule of

thumb, strategy, trick, simplification, or any other kind of
device which drastically limits search for solutions in large
problem spaces.

 Heuristics do not guarantee optimal solutions; in fact, they do

not guarantee any solution at all: all that can be said for a
useful heuristic is that it offers solutions which are good
enough most of the time.
Think like humans
 Cognitive science approach

 Focus not just on behavior and I/O but also look at reasoning
process.

 Computational model should reflect “how” results were

obtained.

 Provide a new language for expressing cognitive theories and

new mechanisms for evaluating them
Think like humans
 GPS (General Problem Solver): Goal not just to produce
humanlike behavior, but to produce a sequence of steps
of the reasoning process that was similar to the steps
followed by a person in solving the same task.

• ELIZA:
A program that
simulated a
psychotherapist
interacting with
a patient and
passed the
Turing Test.
Act like humans

 Behaviorist approach.
 Not interested in how you get results, just the
similarity to what human results are.
 Exemplified by the Turing Test (Alan Turing, 1950).
 Turing Test
 Three rooms contain a person, a computer, and an
interrogator.
Turing test

 A Turing Test is a method of inquiry in artificial intelligence (AI) for determining

whether a computer can think like a human being. 
  The original Turing Test requires three terminals, each of which is physically
separated from the other two. One terminal is operated by a computer, while the
other two are operated by humans.
 During the test, one of the humans functions as the questioner, while the second
human and the computer function as respondents.
 After a preset length of time or number of questions, the questioner is then asked
to decide which respondent was human and which was a computer.
 The test is repeated many times. If the questioner makes the correct determination
in half of the test runs or less, the computer is considered to have artificial
intelligence because the questioner regards it as "just as human" as the human
respondent.
Some Example Applications

 Computer vision: face recognition from a large set

(Interpreting Images)
 Robotics: autonomous (mostly) automobile
 Natural language processing: simple machine translation
 Expert systems: medical diagnosis in a narrow domain
 Spoken language systems: ~1000 word continuous speech
 Planning and scheduling: Hubble Telescope experiments
 Learning: text categorization into ~1000 topics
 Games: Grand Master level in chess (world champion),
checkers, etc.
 Fair Use Notice

The material used in this presentation i.e., pictures/graphs/text, etc.

is solely intended for educational/teaching purpose, offered free of
cost to the students for use under special circumstances of Online
Education due to COVID-19 Lockdown and may include
copyrighted material – the use of which may not have been
specifically authorized by Copyright Owners. It’s application
constitutes Fair Use of any such copyrighted material as provided
in globally accepted law of many countries. The contents of the
presentation are intended only for the attendees of the class being
conducted by the presenter.
Previous Lecture
Overview
 It is the science and engineering of making intelligent machines,
especially intelligent computer programs.
 Two Categories:
• Weak (narrow) AI embodies a system designed to carry out one particular
job.
• Strong AI systems are systems that carry on the tasks considered to be
human-like

 How AI Works:
 Think well
 Act well
 Think like humans
 Act like humans
AI Application Areas
 Game Playing:

 AI plays crucial role in strategic games such as chess, poker, tic-tac-toe, etc., where
machine can think of large number of possible positions based on heuristic
knowledge.

 Involves mostly the research on board games as they have

certain properties that made them ideal subjects for early work
on AI.

 These properties include well defined set of rules, easy board

configuration representation in computers, easy
implementation of state space search and many others.
Search Space – Tic Tac Toe
 Game Playing:
 Games can generate extremely large search spaces.
 These are large and complex enough to require powerful techniques for
determining what alternatives to explore in the problem space.

 These techniques are called heuristics and constitute a major area of AI

research. A heuristic is a useful but potentially fallible problem-solving
strategy:
• Restarting an application when it is not working correctly
• checking to make sure that an unresponsive appliance is plugged in before
assuming that it is broken
• to castle in order to try and protect your queen from capture in a chess game.
 Much of what we commonly call intelligence seems to reside in the
heuristics used by humans to solve problems.
 Game Playing:
 In 2016, DeepMind’s AI-based AlphaGo software defeated Lee
Sedol, the world champion in the game of GO. This is one of the
most significant accomplishments in the field of AI.
 Go has about possible 200 moves per turn, compared to about
20 in Chess. Over the course of a game of Go, there are so
many possible moves that searching through each of them in
advance to identify the best play is too costly from a
computational point of view. Instead, AlphaGo was trained how
to play the game by taking moves played by human experts in
30 million Go games and feeding them into deep-learning neural
networks.
Game Playing:
 In 2011, the computer system IBM Watson made headlines worldwide
when it won the US quiz show Jeopardy!, beating two of the best
players the show had ever produced. To win the show, Watson used
natural language processing and analytics on vast repositories of data
that it processed to answer human-posed questions, often in a fraction
of a second.
 Game Playing
 Shortly after the victory, DeepMind created an advanced version of AlphaGo
called AlphaGo Zero which defeated the predecessor in an AI-AI face off.
• Unlike the original AlphaGo, which DeepMind trained over time by using a
large amount of data and supervision, the advanced system, AlphaGo Zero
taught itself to master the game.
 Other examples of Artificial Intelligence in gaming include the First Encounter
Assault Recon, popularly known as F.E.A.R, which is a first-person shooter
video game.
• The actions taken by the opponent AI are unpredictable because the game is
designed in such a way that the opponents are trained throughout the game
and never repeat the same mistakes. They get better as the game gets
harder. This makes the game very challenging and prompts the players to
constantly switch strategies and never sit in the same position.
Expert Systems:
 These are the programs that serve either in place or for assistance of
human experts. They are interactive and reliable computer-based decision-
making system that rely on the domain knowledge and the problem solving
strategies of human experts incorporated in them. They use both facts and
heuristics to solve complex problems in specified domain.

 DENDRAL: infer the structure of organic molecules from their chemical

formulas.
 MYCIN: used to diagnose bacterial diseases. It served as the base for
establishment of the methodology of development of many other systems.
 PXDES: used to determine the type and level of lung cancer.
 CaDeT: diagnostic support expert system to detect cancer at early stages.
Natural Language Processing
 Natural language processing programs use artificial
intelligence to allow a user to communicate with a computer in
the user's natural language. The computer can both
understand and respond to commands given in a natural
language.

 This involves much more than parsing sentences into

individual parts of speech and looking those words up in a
dictionary.

 It requires extensive
knowledge about the
domain being considered.
Robotics & Planning:
 It is a difficult problem for a number of reasons, not the
least of which is the size of space of possible sequences
of moves.

 For example:
• A simple robot that can move forward, backward, right
or left can have indefinite ways of moving around a
room with obstacles.
• While humans plan effortlessly, creating a computer
program that can do the same is a difficult challenge.
• So this has been an important topic of research for AI
researchers.
 Boston Dynamics :
• dynamic, intelligent and adaptive robots. With nine different
models, the company develops sensor-based controls that
prepare robots for a variety of environments and terrains.

 iRobot
• home-cleaning and lifestyle devices. Its most prominent product,
Roomba is a robot vacuum that maps and adapts to its
environment as it clears dirt and small debris from floors.
 Neurala
• The Neurala Brain,” AI software that makes an array of devices
more intelligent. Already integrated in more than nine million
devices, the technology boosts intelligence in cars, phones, drones
and cameras. It's also used by major organizations like NASA,
DARPA, Motorola and NVIDIA.

 Hanson Robotics:
• human-like robots that not only have human appearances but also
characteristics like eye contact, facial recognition, speech and the
ability to hold natural conversations.
• well-known for its highly conversational humanoid robot,
Sophia, featured in countless news segments,
discussion panels and technology conferences.
Tapia
Aden

iCub

Asimo Sophia
Machine Learning:
 Machine-learning algorithms use statistics to find patterns in
massive amounts of data. And data, here, encompasses a lot of
things: numbers, words, images, clicks. If it can be digitally
stored, it can be fed into a machine-learning algorithm.

 Machine learning is the process that powers many of the

services we use today:
• recommendation systems like those on Netflix, YouTube, and
Spotify
• search engines like Google and Baidu
• social-media feeds like Facebook and Twitter
• voice assistants like Siri and Alexa.
Marketing
 Artificially intelligent systems constantly work on the background of
popular products and services such as Netflix, Amazon, and,
naturally, Google.
 AI in marketing is the use of customer data, machine learning and
other computational concepts to predict a person’s action or inaction.
It can take on huge amounts of data and help marketers easily
segment them. As such, marketers can further break down data to
create customized content for their audiences.
 Understand Customer
 Find the Trend
 Content Generation
 Dynamic Pricing
Artificial Neural Networks:
 An artificial neural network (ANN), usually called "neural
network" (NN), is a mathematical model or computational
model that tries to simulate the structure and/or functional
aspects of biological neural networks. It consists of an
interconnected group of artificial neurons and processes
information using a connectionist approach to computation.
Computer Vision:
 Computer vision is the science and technology of machines that see, where see in this
case means that the machine is able to extract information from an image that is
necessary to solve some task.

 Computer vision is concerned with the theory behind artificial systems that extract
information from images. The image data can take many forms, such as video sequences,
views from multiple cameras, or multi-dimensional data from a medical scanner.

 The goal of computer vision research is to give computers the same powerful facility for
understanding of their surroundings.
Space Exploration

 Space expeditions and discoveries always require analyzing vast

amounts of data. Artificial Intelligence and Machine learning is
the best way to handle and process data on this scale. After
rigorous research, astronomers used Artificial Intelligence to sift
through years of data obtained by the Kepler telescope in order
to identify a distant eight-planet solar system.

Artificial Intelligence is also being

used for NASA’s next rover mission to
Mars, the Mars 2020 Rover. The
AEGIS, which is an AI-based Mars
rover is already on the red planet. The
rover is responsible for autonomous
targeting of cameras in order to
perform investigations on Mars.
Autonomous Vehicles
 In 2009 Google showed it was possible for its self-driving Toyota Prius to complete
more than 10 journeys of 100 miles each, setting society on a path towards
driverless vehicles.
 Companies like Waymo conducted several test drives in Phoenix before deploying
their first AI-based public ride-hailing service. The AI system collects data from the
vehicles radar, cameras, GPS, and cloud services to produce control signals that
operate the vehicle.
 Advanced Deep Learning algorithms can accurately predict what objects in the
vehicle’s vicinity are likely to do. This makes Waymo cars more effective and safer.
 Another famous example of an autonomous vehicle is Tesla’s self-driving car.
Artificial Intelligence implements computer vision, image detection and deep
learning to build cars that can automatically detect objects and drive around without
human intervention.
Automated Reasoning and Theorem proving:
 When the system is required to do something that it has not
been explicitly told how to do, it must reason - it must figure
out what it needs to know from what it already knows.

 The study in automated reasoning helps produce software

which allows computers to reason completely or nearly
completely, automatically.

 Automated theorem proving in general attempts to find proofs

to theorems which are usually assumed to be true.
 Early efforts to create theorem provers were not much
successful because of generation of infinite non-relevant
provable theorems. But as the time grew, the results got better
and better.
 Languages & Environments:
 Major AI languages are Python, R and Java.

 PYTHON is a favorite choice in the list of all AI development languages

due to the simplicity.
 Python is very simple and can be easily learnt.
 It takes short development time in comparison to other languages like
Java, C++ or Ruby.

 R is the most effective language and environment for analyzing and

manipulating data for statistical purposes.
 Using R, we can easily produce well-designed publication-quality plots,
including mathematical symbols and formulae where needed.
 It has numerous packages like RODBC, Gmodels, Class and Tm which
are used in the field of machine learning.