Progress Report – July 2009

In this progress report, I attended the ICAI 09 conference in Las Vegas (the IJCAI 2009 is located in Pasadena).  I presented my work on A Brute Force Approach to Solving the Knights Tour Using Prolog.  One thing that I learned is the participants are proud to share their work with others.  Another lesson is to make your presentation within the time allotted.  A few presenters did not finish their presentations in the 20 minutes time slot.  I rehearsed my presentation a few times to have the timing within 20 minutes.  I also learned that I need to get more training with MS Power Point.

I am also continue to view the video lectures from the Summer Schools in Logic and Learning from Video Lectures.  Thus, I have completed the Intelligent Agents video lectures by John Lloyd. 

AIMA Part VI

Part VI of the AIMA book covers learning and various techniques on agents can learn.  The material covers Learning from Observations, Neural Networks, Reinforcement Learning, and Knowledge in Learning.  In the second edition, the chapter in Neural Networks is replaced with Statistical Learning Methods.

I read Chapter 18 of the AIMA book called Learning from Observations.  The Chapter focuses on decision trees and decision lists as some computational learning theory.  The main classification algorithm is the ID3 algorithm for classifying examples and generating decision trees.  The ID3 algorithm is based on Hunt’s Concept Learning System (CLS).  The ID3 algorithm uses information theory for obtaining the decision tree covering most the examples.  I have studied the ID3 algorithm quite extensively since the original application was to classify the winning position in the king and rook versus the king and knight endgame (Quinlan 1983).

Progress Report – May 2009

This month I submitted my paper A Brute Force Approach to Solving the Knight’s Tour using Prolog to the ICAI ‘09 conference, located at the Monte Carlo in Las Vegas, Nevada.  According to the conference chair, the conference acceptance rate was 27 percent.  This conference is in conflict the IJCAI conference in Pasadena, California.

Nevertheless, besides my research paper, much of my activity this month has been around logic programming and inductive logic programming (ILP).  I ordered from Amazon several books recently around this area, in particular The Art of Prolog and Inductive Logic Programming: From Machine Learning to Software Engineering.  The Art of Prolog has a gentle introduction to logic programming – it is very clear and instructive.  From this book I was able to understand early articles in Logic Programming.

Progress Report – Q1 2009

This is another progress report of my AI journey.  The top posts for the past quarter as follows:

• Wumpus World
• Stanford CS229 Class
• About
• What is Machine Learning?
• AIMA Part IV Acting Logically
• Progress Report – December 2008
• Logic, Programming and Prolog

AIMA Part IV Acting Logically

Continuing my quest to read AI: Modern Approach (Russell & Norvig 1995), Part IV is the section on Acting Logically. Chapter 11 starts with a planning agent, and the authors develop the partial order planner (POP) algorithm. Please note the planning agent is not a theorem prover or a problem solving agent, but a specialized agent with a directed search in a planning space. The authors continue with the STRIPS language to develop the operators for a planning agent.

Chapter 12 starts with the discussion of current planners (circa 1995) and its use in industry and governement.  The planning topic continues with the hierarchical decomposition, generating the HD-POP algorithm.  The authors continue with conditional effects and universal quantification, leading to the  POP-DUNC algorithm.  Finally actions have resource constraints such as materials, cost, and time.

Lastly, Chapter 13 begins a discussion regarding conditional or contingency planning.  The algorithm for a conditional planning agent is presented as well as the algorithm for Conditional Partial Order Planner (CPOP).  Next is the section on a simple replanning agent, and an algorithm for a replanning agent is presented.  The algorithm for a situated planning agent is presented.  In summary, planning agents need to account for real world issues such as incomplete information, execution errors, or unmet preconditions and have contingency plans, quickly replan, or adjust the plan as it monitors the situation during the execution.

AI planners have been used in industry for job scheduling and by governments for job scheduling for the Hubble Telescope by NASA as well as for space missions by the ESA.  Other planners are discussed by history remarks in all three chapters.

The last two sections conclude knowledge and reasoning along with acting logically.  The basis is first order logic (FOL) along with situational calculus.  These topics work well with logic programming and the prolog programming language.

AIMA Part III Knowledge and Reasoning

I just completed reading the last two chapters (9 & 10) of Part III, called Knowledge and Reasoning, of AI: Modern Approach (Russell & Norvig 1995).  Basically, Chapter 9 wraps up the previous material in Chapters 6 (Propositional Logic) and 7 (Predicate Calculus) by defining a generalized Modus Ponens and linked that Modus Ponens is a subset of the Resolution Refutation principle (Robinson 1965).  Resolution Refutation principle can be used to prove all statements in first order logic.  Also, Chapter 9 illustrates the Forward-Chaining, Backward-Chaining, Unification algorithms.  Finally, Chapter 10 discusses indexing knowledge bases, theorem provers, logic programming, productions systems, semantic networks, and meta systems.  The authors discuss the advantages and disadvantages of each systems and its primary application uses.  In both Chapters, the historical remarks are very informative along with relavent references.