Artificial Intelligence

CSE 5290/4301
Florida Institute of Technology
Instructor: Debasis Mitra


Department: Computer Sciences

Abstract

Artificial Intelligence or AI possibly generates more curiosity than any other branches in computer science. In a sense, AI is THE purpose of computer science! Why? Because, "computing" or doing arithmetic was one of the foremost "intelligent" behaviors that was dreamt to be mechanized. When it became clear that many other human activities may be mapped to arithmetic, which the early computers were so efficient with, the term AI was coined in 1956. Since then AI has shaped up as a sub-topic within CS. In this course we will make a short journey through those materials. After finishing this course, hopefully, you will be (1) knowledgeable enough to learn more, if and when you need it, (2) join the workforce where some form of AI is deployed, and (3) demystify AI to your friends and relatives!

In 2023, a few more words on "demystifying" AI is needed! AI has a sub-topic called Machine Learning (ML, we will cover its basics). One of the algorithms for ML uses Artificial Neural Networks (ANN, inspired by human brain). Some ANN's with deep layers are showing miraculous success in real life over the last few years, to the extent that the Deep Learning is replacing many tasks traditionally performed by the CS algorithms (e.g., natural language generation), and performing some other tasks that were dreams for automation (e.g., self-driving cars). This has created a myth that "AI" is same as "Deep Learning." So be it, but you should be wiser!

The CSE4301 syllabus is: here, and the CSE 5290 syllabus is here

The Graduate Comprehensive Exam's syllabus, Some topics for Graduate comprehensive exam may not be covered in the class.
A few samples: CompsAiSpr17, CompsAiSpr18,

Florida Tech Students' Handbook on Cheating and Plagiarism: http://www.fit.edu/studenthandbook/print.php#policy_2490

Prerequisites:
(1) Discrete Mathematics, (2) Data structures and algorithms, and (3) Programming in a higher level language.

A historical perspective on Artificial Intelligence

===================== Spring 2024 =================
Spring 2024 CLASS coordinates: 3:30 - 4:45 pm TR Olin Life Sciences 129

Continuously updated: day to day schedule

>>>>>>>>>>>>>>>>>>>>>>>>>>>> WARNING:
Disclaimers to the lecture notes:
THESE NOTES ARE FOR HELPING YOU TO STUDY THE TEXT BOOK. I KEEP UPDATING THESE AND WILL NOT BE RESPONSIBLE IF YOU FIND THAT THE NOTES HAVE CHANGED AFTER YOU HAVE LAST VIEWED IT. ON ANY CONTENT CONFLICTING WITH TEXTBOOK, TEXTBOOK WINS BY DEFAULT. HOWEVER, PLEASE BRING THAT TO MY NOTICE. TEXT MAY HAVE OLD INFORMATION OR DIFFERENT VIEWPOINT THAN THAT OF MINE.

MODULE TOPIC (MySlides) TEXT Slides (3rd ed)
Background Introduction on AI and Background in CS Read early chapters
SEARCH Problem Solving with Search, chapter04a.pdf,
Local Search algorithms chapter04b.pdf
Adversarial Game Search, chapter06.pdf
Constraints Reasoning, chapter05.pdf
========================== ======================================= =======================================
LOGIC Automated Reasoning with Propositional Logic chapter07.pdf,
Additional:
DPLL
Automated Reasoning with Predicate Logic chapter08.pdf,
Inferencing:
chapter09.pdf ,
Additional: Problem with Existential Quantifier with implication,
Example of Skolemization and Existential Quantifier elimination,
Second example of resolution in Predicate logic
Additional materials on logic
Sample Knowledge base on a book page
Sample Expert system code with CLIPS
Sample Prolog code
Sample automated reasoning code with Otter system
None
========================== ======================================= =======================================
UNCERTAINTY Probabilistic Reasoning chapter13.pdf
Bayesian Network chapter14a.pdf
Bayesian Inferencing chapter14b.pdf (up to slide 10)
========================== ======================================= =======================================
MACHINE LEARNING Basics + Decision Tree chapter18.pdf
Classification, Regression, Unsupervised learning, etc. 18LearningSlides.pdf, 18LearningSlides.pptx
Machine Learning with Neural Networks chapter20b.pdf
========================== ======================================= =======================================
THOUGHTS ON AI & ETHICS Problems to watch on AIethics.ppt
========================== ======================================= =======================================

-------------- Resources:
A recent (May'19) survey from MIT-Tech Reviews finds a slowing-down trend for deep learning . A news item on US interest, 2019.
A discrete math online book from Dartmout: https://math.dartmouth.edu/archive/m19w03/public_html/book.html
Artificial Intelligence IS Computer Science! Turing's Imitation game-paper, 1950 (20pg).
First workshop Proposing the term AI, 1955.
Open letter on AI by Stephen Hawking, Elen Musk and others.
A paper on AI and ethics by Bostrom-Yudkowsky (21 pages).
US Govt. Strategic Plan, 2016.
Hidden Markov Model: a conoical tutorial by Lawrence Rabiner, 1988 (30pg).
Nice set of Data Science / ML interview questions

+++++++++++++++++++ Lecture Notes +++++++++++++
Text book: S. Russell and P. Norvig, Artificial Intelligence: Modern Approach. Pearson, third ed., 2010. http://aima.cs.berkeley.edu/

Figures
Textbook slides , change chapter number on url

Introduction on AI and Background in CS
Complexity theory-lite (7% with surgeon general's warning, etc.)
Problem Solving with Search, and Search algorithms in Text Ch-4, Local Search algorithms in Text ,
Game Search:
My Notes, and from the text.
A background material on NP-completeness
Reasoning with Constraints, and from the text Ch-5, examples from Dechter's book, a PC example , and constraints counter examples from Edward's book,
A bit more animated lecture slides on constraints,
Another movie animation on some constraint reasoning algorithms from Andrew Moore at CMU.
Spatio-temporal constraints

Automated Reasoning with Propositional Logic Ch-7,
DPLL
Automated Reasoning with Predicate Logic, Problem with Existential Quantifier with implication,
Example of Skolemization and Existential Quantifier elimination,
Second example of resolution in Predicate logic
Sample Knowledge base on a book page
Inferencing in Predicate Logic
Sample Expert system code with CLIPS
Sample Prolog code
Sample automated reasoning code with Otter system

Modeling Uncertainty
I am jotting down some additional clarification
notes on probabilistic reasoning.
Reasoning with Uncertainty, Part-II: Inferencing
Temporal probabilistic network, Part-II (we will not cover these two parts)

Machine learning-I
More on learning , from Ch 18 (my text notes 18.6 onwards)
Machine learning with neural networks
My slides on Machine learning (additional materials to above)

AI and ethics: My thoughts, read also Chapter 26.3, 3rd ed. AIMA text


Statistical Machine learning I
Classical Planning

Materials are copyrighted to me (year 2019). Many materials were developed before I joined FIT. E-mail: dmitra at cs.fit.edu