ReadingNotes

Artificial Intelligence A Modern Approach

Artificial Intelligence: A Modern Approach by Stuart Russell and Peter Norvig - ISBN 0136042597 - Prentice Hall 2009 (3rd Ed.)

Motivation

Recommended by several persons including J-M Vanel, also required for AIClass.

Pre-reading model

List the main decisions that should be impacted by reading the book in order to estimate the Seedea:Content/Newconcepts#InformationActionRatio and facilitate rule design as suggested after finishing TheTinkerersAccomplice.

Draw a schema (using PmGraphViz or another solution) of the situation of the area in the studied domain before having read the book.

Reading

Part I Artificial Intelligence

1 Introduction
- add the seminal Lisp article I read in Berkeley, important for the history of AI
- brief mention of AGI
  - see Peter Norvig's most recent AGI 2011: Thursday Evening Opening Remarks with Moshe Looks, August 2011
2 Intelligent Agents

Part II Problem Solving

3 Solving Problems by Searching
4 Beyond Classical Search
5 Adversarial Search
- see GPP
6 Constraint Satisfaction Problems

Part III Knowledge and Reasoning

7 Logical Agents
8 First-Order Logic
9 Inference in First-Order Logic
10 Classical Planning
- motivated by discussions on with Beniz and GPP (cf StrategyLessons) then later on WithoutNotesSeptember11#EmergenceOfCreativityASimulationApproach and more generally my OwnConcepts#UniversalComposer
- important discussion on complexity and its implication
  - see also http://qwiki.stanford.edu/index.php/Complexity_Zoo:P#pspace
- point of PDDL expressiveness and its limits
  - overall limits of expressiveness should be considered for Heuristics#ExpressivePower
- see also Wikipedia:Automated planning and scheduling, Wikipedia:Planning Domain Definition Language (PDDL), Wikipedia:Satplan and Scholarpedia:Action selection
- see I:BooksExercises/AIMA#Chapter10
- Goal-Oriented Action Planning (GOAP) as used in F.E.A.R.
11 Planning and Acting in the Real World
- same motivation than from the previous chapter
12 Knowledge Representation

Part IV Uncertain Knowledge and Reasoning

13 Quantifying Uncertainty
14 Probabilistic Reasoning
15 Probabilistic Reasoning over Time
16 Making Simple Decisions
17 Making Complex Decisions

Part V Learning

18 Learning from Examples
- Wikipedia:Computational learning theory
- Wikipedia:Probably approximately correct learning aka PAC learning
  - Valiant's work discovered after WithoutNotesAugust11#ScottAaronson
- to avoid overfitting or Ockham's razor principle mention of Wikipedia:Minimum description length aka MDL and Wikipedia:Kolmogorov complexity in the historical notes
- Wikipedia:Locality-sensitive hashing aka LSH, to explore as a link with metics thus FromGeometryToTopology
19 Knowledge in Learning
20 Learning Probabilistic Models
21 Reinforcement Learning
- TD learning discovered before via WithoutNotesSeptember10#JonahLehrer
- active vs. passive learning agent sounds coherent with constructivist principles
- Wikipedia:Q-learning action/value pair would also work for self-model discovery via correlation as proposed in TreeOfKnowledge#SelfUnderstandingPartOfHomeostase

Part VII Communicating, Perceiving, and Acting

22 Natural Language Processing
23 Natural Language for Communication
24 Perception
25 Robotics

Part VIII Conclusions

26 Philosophical Foundations
27 AI: The Present and Future

Overall remarks and questions

this? that?

Synthesis

So in the end, it was about X and was based on Y.

Critics

Point A, B and C are debatable because of e, f and j.

Vocabulary

(:new_vocabulary_start:) new_word (:new_vocabulary_end:)

Post-reading model

Draw a schema (using PmGraphViz or another solution) of the situation of the area in the studied domain after having read the book. Link it to the pre-reading model and align the two to help easy comparison.

Fabien Benetou's PIM