Instructor Prof. Gwo-Boa Horng
Lectures Time: Wed. 1:10 - 4:00pm;
Place: Room 701
Textbook Artificial Intelligence (2nd ed.), E. Rich and K. Knight,

 

Date Lecture Notes Homework Notation
9/12

Introduction :

  * Course overview

  * AI history

  * The underlying assumption

  * Turing test

  * The challenge of AI  

   
9/19

AI techniques:

  * search (game, playing)

  * use of knowledge (question answering)

  * abstraction (pattern recognition)

State Space Search Problems

  * definition

  * example : water jug problem

7 problem characteristics 

   
9/26

   * Problem characteristics (cont.)

   * BFS and DFS

   * LISP

下載:第一次作業   
10/3

  * LISP (cont.)

  * Heuristic search techniques

      --generate-and-test

      -- hill climbing

      -- simulated annealing

      -- best-first

第二次作業 : (due 10.17) Use LISP to implement tic-tac-toe program 2 (pp. 10-11)  
10/10 國慶日    
10/17

  * Heuristic Search Techniques (cont.)

      --best first search

            * A* algorithm

            * heuristic functions

      -- agenda-driven search

      -- problem reduction

            * AND/OR graphs

      -- Constraint satisfaction

            * waltz algorithm

      -- means-ends analysis                                  

   
10/24

  * Game Playing

      -- minimax search

      -- alpha-beta cutoffs

      -- iterative deepening

  * Knowledge Representation

      -- introduction 

Programming Project (I): 

1. problem formulation. 

2. Heuristic search.(完成日:一個月後 ) 

 

 

 
10/31

  * knowledge representation

      -- introduction

      -- approaches to K.R.

      -- issues in K.R.    

      -- the frame problem 

#3.3

#3.5

#3.14

#12.1

#12.2

#14.2

 
11/7

  * Representing knowledge using rules

      -- forward rules, backward rules

      -- advantages and disadvantages

      -- rule-based reasoning system

       * control strategy

  * Expert systems

      -- definition

      -- comparison with conventional systems

      -- structure of E.S.

   
11/14 期中考    
11/21

  * Structured knowledge representations

      -- weak slot-and-filler structures

           * semantic nets

           * frames

      -- strong slot-and-filler structures

           * conceptual dependency

           * scripts

   
11/28

  * Natural language processing (NLP)

      -- motivations

      -- why NLP is hard

      -- NLP tasks

      --overview the steps

      -- syntactic processing 

Programming Project (II):

Natural Language Interface 

(due 12/19)

 
12/5

  * NLP (cont.)

      -- semantic analysis

      -- discourse integration

      -- pragmatic analysis

      -- conclusions

   
12/12

  * Learning

      -- what is "learning"

      -- approaches in this area

      -- generic learning model

      -- learning strategies

          * rote learning

          * learning by taking advice

          * learning in problem solving

          * learning from example 

Programming Project (III):

Learning (due 1/9)

 
12/19

  * Learning (cont.)

      -- Learning from examples

      -- Explanation-based learning

      -- Discovery

      -- Analogy

   
12/26

  * Statistical reasoning

      -- Probability

      -- Bayesian networks

      -- Dempster-Shafer theory

      --Ad hoc methods 

   
1/2   * Project presentation     
1/9 期末考    

 

References:
* R. Reddy, "The challenge of artificial intelligence", IEEE Computer Oct. 1996
* M. Krol, "Have we witnessed a real life Turing test?" IEEE Computer Mar. 1999
* G. Buttazzo, "Artificial consciousness : Utopia or real possibility?", IEEE Computer July 2001