CmpE
530 Mathematical Fundamentals of Artificial Intelligence
Fall 2004
Fikret
Gurgen
Class
time: M 234 Off hours:
M 1 Tue 23 and any time available
The goal of Artificial Intelligence (AI) has always been
to imitate human intelligence but the methodology has changed radically.
Previously learning was not that important but now learning is the centerpiece
of AI research. Now human brain function is seeing major emphasis. Previously
models of behavior were exclusively symbolic, but now sub-symbolic models
receive the major focus. As the result of all these trends is that AI
researchers need a background in mathematically oriented subjects such as
probability theory, information theory, optimization theory, etc. The purpose of
this course is to introduce certain background in order to describe
computational models of intelligent behavior and their relationship to
structures in the brain.
TEXT
MATERIAL:
-
Introduction
to Applied Statistical Signal Analysis, Second
Ed., Richard Shiavi, Academic Press, 1999.
References:
“Digital Signal Processing
Fundamentals”, “Probability and Statistics”, “Fuzzy Theory” and “Information
Theory” text examples:
Digital Signal Processing, Rabiner et al., Proakis et al.
(nice signal processing!),
Probability and Statistics, by M. DeGroot. (a good old one!),
Fuzzy Theory, by Bart Kosko (a fuzzy theory
book),
Statistical Pattern Recognition by Fukunaga (classical
book)
Any Artificial
Intelligence book
RECOMMENDED
TOOLS:
1) Textbooks
ready programs by CD
2) MATLAB, C/C++
programming
TENTATIVE
TOPICS:
1.
Introduction to signal terminology and signal
processing
1.1. Domain Types
1.2. Fourier Analysis and sampled-data
signals
1.3. Continuous-time and Discrete-time
system analysis
2. Empirical
modelling and approximation
2.1. Generalized least
squares
2.2. Interpolation and
extrapolation
3. Learning
Fundamentals (notes)
3.1. Fundamental
concepts
4.
Probability concept and signal characteristrics
4.1.
Distributions
4.2.
Characteristics
5.
Introduction to random processes and signal
correlation
5.1. Stationarity and moment
functions
5.2. Correlation and signal
structure
5.3. Markov chain and hidden
markov model (HMM)
6.
Information Theory
6.1. Definitions, entropy, mutual
information
7. Fuzzy
concept and applications
7.1 Fuzziness and
probability
7.2 Fuzzy definitions and
examples
8. Rule
based applications
8.1 Rule definitions and
examples
COURSE
GRADING:
Midterm % 25
Presentation+
Project + Homeworks + Quizzes % 40
Final % 35
IMPORTANT
RULES: