Analytical Learning.

Slides:



Advertisements
Similar presentations
Computational Learning Theory
Advertisements

Machine learning Overview
Artificial Neural Networks
Explanation-Based Learning (borrowed from mooney et al)
Concept Learning and the General-to-Specific Ordering
Combining Inductive and Analytical Learning
ARTIFICIAL INTELLIGENCE [INTELLIGENT AGENTS PARADIGM] Professor Janis Grundspenkis Riga Technical University Faculty of Computer Science and Information.
Computer Science CPSC 322 Lecture 25 Top Down Proof Procedure (Ch 5.2.2)
Naïve Bayes. Bayesian Reasoning Bayesian reasoning provides a probabilistic approach to inference. It is based on the assumption that the quantities of.
 2002, G.Tecuci, Learning Agents Laboratory 1 Learning Agents Laboratory Computer Science Department George Mason University Prof. Gheorghe Tecuci Deductive.
CHAPTER 13 Inference Techniques. Reasoning in Artificial Intelligence n Knowledge must be processed (reasoned with) n Computer program accesses knowledge.
Knowledge Representation and Reasoning Learning Sets of Rules and Analytical Learning Harris Georgiou – 4.
Combining Inductive and Analytical Learning Ch 12. in Machine Learning Tom M. Mitchell 고려대학교 자연어처리 연구실 한 경 수
Knowledge in Learning Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 19 Spring 2004.
처음 페이지로 이동 Chapter 11: Analytical Learning Inductive learning training examples n Analytical learning prior knowledge + deductive reasoning n Explanation.
Learning control knowledge and case-based planning Jim Blythe, with additional slides from presentations by Manuela Veloso.
Decision Tree Learning
Knowledge in Learning Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 19 Spring 2005.
LEARNING FROM OBSERVATIONS Yılmaz KILIÇASLAN. Definition Learning takes place as the agent observes its interactions with the world and its own decision-making.
Relational Data Mining in Finance Haonan Zhang CFWin /04/2003.
November 10, 2009Introduction to Cognitive Science Lecture 17: Game-Playing Algorithms 1 Decision Trees Many classes of problems can be formalized as search.
Instance-Based Learning
Data Mining with Decision Trees Lutz Hamel Dept. of Computer Science and Statistics University of Rhode Island.
Machine Learning CSE 473. © Daniel S. Weld Topics Agency Problem Spaces Search Knowledge Representation Reinforcement Learning InferencePlanning.
LEARNING FROM OBSERVATIONS Yılmaz KILIÇASLAN. Definition Learning takes place as the agent observes its interactions with the world and its own decision-making.
Logical Agents Chapter 7 Feb 26, Knowledge and Reasoning Knowledge of action outcome enables problem solving –a reflex agent can only find way from.
Semantics with Applications Mooly Sagiv Schrirber html:// Textbooks:Winskel The.
Describing Syntax and Semantics
Lehrstuhl für Informatik 2 Gabriella Kókai: Maschine Learning 1 Evaluating Hypotheses.
CSCI 5582 Fall 2006 CSCI 5582 Artificial Intelligence Lecture 22 Jim Martin.
MACHINE LEARNING. What is learning? A computer program learns if it improves its performance at some task through experience (T. Mitchell, 1997) A computer.
17.5 Rule Learning Given the importance of rule-based systems and the human effort that is required to elicit good rules from experts, it is natural to.
Inductive Logic Programming Includes slides by Luis Tari CS7741L16ILP.
Notes for Chapter 12 Logic Programming The AI War Basic Concepts of Logic Programming Prolog Review questions.
1 Machine Learning: Lecture 11 Analytical Learning / Explanation-Based Learning (Based on Chapter 11 of Mitchell, T., Machine Learning, 1997)
Machine Learning Chapter 11. Analytical Learning
CpSc 810: Machine Learning Design a learning system.
1 Machine Learning What is learning?. 2 Machine Learning What is learning? “That is what learning is. You suddenly understand something you've understood.
Kansas State University Department of Computing and Information Sciences CIS 830: Advanced Topics in Artificial Intelligence Friday, February 4, 2000 Lijun.
November 10, Machine Learning: Lecture 9 Rule Learning / Inductive Logic Programming.
Machine Learning Chapter 2. Concept Learning and The General-to-specific Ordering Tom M. Mitchell.
Multi-Relational Data Mining: An Introduction Joe Paulowskey.
Machine Learning Chapter 5. Artificial IntelligenceChapter 52 Learning 1. Rote learning rote( โรท ) n. วิถีทาง, ทางเดิน, วิธีการตามปกติ, (by rote จากความทรงจำ.
Outline Inductive bias General-to specific ordering of hypotheses
Overview Concept Learning Representation Inductive Learning Hypothesis
For Monday Finish chapter 19 No homework. Program 4 Any questions?
Ch. 13 Ch. 131 jcmt CSE 3302 Programming Languages CSE3302 Programming Languages (notes?) Dr. Carter Tiernan.
For Monday Finish chapter 19 Take-home exam due. Program 4 Any questions?
CS 5751 Machine Learning Chapter 10 Learning Sets of Rules1 Learning Sets of Rules Sequential covering algorithms FOIL Induction as the inverse of deduction.
© Copyright 2008 STI INNSBRUCK Intelligent Systems Propositional Logic.
Machine Learning Concept Learning General-to Specific Ordering
CpSc 810: Machine Learning Analytical learning. 2 Copy Right Notice Most slides in this presentation are adopted from slides of text book and various.
Data Mining and Decision Support
1 Propositional Logic Limits The expressive power of propositional logic is limited. The assumption is that everything can be expressed by simple facts.
CS 5751 Machine Learning Chapter 12 Comb. Inductive/Analytical 1 Combining Inductive and Analytical Learning Why combine inductive and analytical learning?
Logical Agents Chapter 7. Outline Knowledge-based agents Propositional (Boolean) logic Equivalence, validity, satisfiability Inference rules and theorem.
More Symbolic Learning CPSC 386 Artificial Intelligence Ellen Walker Hiram College.
Concept Learning and The General-To Specific Ordering
Computational Learning Theory Part 1: Preliminaries 1.
Chap. 10 Learning Sets of Rules 박성배 서울대학교 컴퓨터공학과.
Anifuddin Azis LEARNING. Why is learning important? So far we have assumed we know how the world works Rules of queens puzzle Rules of chess Knowledge.
Logical Agents. Outline Knowledge-based agents Logic in general - models and entailment Propositional (Boolean) logic Equivalence, validity, satisfiability.
Writing a sound proposal
CS 9633 Machine Learning Explanation Based Learning
CS 9633 Machine Learning Concept Learning
CS 9633 Machine Learning Inductive-Analytical Methods
Data Mining Lecture 11.
Knowledge in Learning Chapter 19
This Lecture Substitution model
Lecture 14 Learning Inductive inference
Presentation transcript:

Analytical Learning

Content Introduction Learning with Perfect Domain Theories: PROLOG-EBG Difference between Inductive and Analytical Learning Learning with Perfect Domain Theories: PROLOG-EBG Remarks on Explanation-Based Learning Summary

Introduction Explanation is used to distinguish the relevant features of the training examples from the irrelevant ones, so that the examples can be generalised Prior knowledge and deductive reasoning is used to augment the information provided by the training examples Prior knowledge is used to reduce the complexity of hypothesis space Assumption: learner's prior knowledge is correct and complete

Introduction 2 Example: Learn to recognise important classes of games Goal: Recognise chessboard positions in which black will lose its queen within two moves Induction can be employed <=> Problem: thousands of training examples similar to this one are needed Suggested target hypothesis: board position in which the black king and queen are simultaneously attacked Not suggested: board position in which four white pawns are still in their original location

Introduction 3 Explanations of human beings provide the information needed to rationally generalise from details Prior knowledge: e.g. knowledge about the rules of chess: legal moves, how is the game won, ... Given just this prior knowledge it is possible in principle to calculate the optimal chess move for any board position <=> in practice it will be frustratingly complex Goal: learning algorithm that automatically constructs and learns a move from such explanations

Difference between Inductive and Analytical Learning Analytical learning methods seek a hypothesis that fits the learner's prior knowledge and covers the training examples Explanation based learning is a form of analytical learning in which the learner processes each new training example by Explaining the observed target value for this example in terms of the domain theory Analysing this explanation to determine the general conditions under which the explanation holds Refining its hypothesis to incorporate these general conditions

Difference between Inductive and Analytical Learning (2) Difference: They assume two different formulations of the learning problem: Inductive learning: input: hypothesis space H + set of training examples output: hypothesis h, that is consistent with these training examples Analytical learning: input: hypothesis space H + set of training examples + domain theory B consisting of background knowledge (used to explain the training examples) output: hypothesis h, that is consistent with both the training examples D and the domain theory B

Difference between Inductive and Analytical Learning (3) Illustration: is True if is a situation in which black will lose its queen within two moves and False otherwise H: set of Horn-clauses where predicates used by the rules refer to the position or relative position of specific pieces B: formalisation of the rules of chess

New Example Given: Target concept: Instance space X: Each instance describes a pair of objects represented by the predicates Type, Color, Volume, Owner, Material, Density and On. Hypothesis space H: Each hypothesis is a set of Horn clauses. The head of each clause is a literal of the target predicate SafeToStack. The body of each Horn clause is a conjunction of literals based on the same predicates used to describe the instances + LessThan|Equal|GreaterThan + function: plus|minus|times Target concept: Training examples: On(Obj1, Obj2) Owner (Obj1, Fred) Type(Obj1, Box) Owner (Obj2, Louise) Type(Obj2, Endtable) Density(Obj1, 0.3) Color(Obj1, red) Material (Obj1, Carboard) Color(Obj2, Blue) Material (Obj1, Wood) Volume(Obj1, 2)

New Example 2 Determine: Domain Theory B: A hypothesis from H consistent with the training examples and the domain theory

Content Introduction Learning with Perfect Domain Theories: PROLOG-EBG An Illustrative Trace Remarks on Explanation-Based Learning Explanation-Based Learning of Search Control Knowledge Summary

Learning with Perfect Domain Theories: PROLOG-EBG A domain theory is said to be correct if each of its assertions is a truthful statement about the world A domain theory is complete with respect to a given target concept and instance space, if the domain theory covers every positive example in the instance space. It is not required that the domain theory is able to prove that negative examples do not satisfy the target concept.

Learning with Perfect Domain Theories: PROLOG-EBG (2) Question: The learner had a perfect domain theory, why would it need to learn? Answer: There are cases in which it is feasible to provide a perfect domain theory It is unreasonable to assure that a perfect domain theory is available. A realistic assumption is that plausible explanations based on imperfect domain theories must be used, rather than exact proofs based on perfect knowledge.

Learning with Perfect Domain Theories: PROLOG-EBG (3) PROLOG-EBG (Kedar-Cabelli and McCarthy 1987) Sequential covering algorithm When given a complete and correct domain theory, the method is guaranteed to output a hypothesis (set of rules) that is correct and that covers the observed positive training examples Output: set of logically sufficient conditions for the target concept, according the domain theory

Learning with Perfect Domain Theories: PROLOG-EBG (4) Repeatedly: Domain theory is correct and complete this explanation constitutes a proof that the training examples satisfy the target concept PROLOG-EBG(TargetConcept, TrainingExamples, Domain Theory) LearnedRules Pos the positive examples from TrainingExamples for each PositiveExample in Pos that is not covered by LearnedRules do Explain Explanation an explanation (proof) in terms of the DomainTheory that PositiveExample satisfies the TargetConcept Analyse Sufficient Condition the most general set of features of PositiveExample sufficient to satisfy the TargetConcept according to the Explanation Refine LearnedRules LearnedRules + NewHornClause, where NewHornClause is of the form TargetConcept SufficientConditions In general there may be multiple possible explanations Any or all of the explanations may be used. Explanation is generated using backward chaining search as performed by PROLOG.

An Illustrative Trace (2) The imprtant question in the generalising-process: Of the many features that happen to be true of the current training example, which ones are generally relevant to the target concept? Explanation constructs the answer: Precisely the features mentioned in the explanation

An Illustrative Trace (3) General rule justified by the domain theory: leaf node in the proof tree expects Equal(0.6,times(2,03) and LessThan(0.6,5)

An Illustrative Trace (4) Explanation of the training example forms the proof for the correctness of this rule PROLOG-EBG computes the most general rule that can be justified by the explanation, by computing the weakest preimage of the explanation Definition: The weakest preimage of a conclusion C with respect to a proof P is the most general set of initial assertions A, such that A entails C according to P. Example: PROLOG_EBG computes the weakest preimage of the target concept with respect to the explanation, using a general procedure called regression Regression: go iteratively backward through the explanation, first computing the weakest preimage of the target concept with respect to the final proof step in the explanation Computing the weakest preimage of the resulting expressions with respect to the proceeding step and so on

An Illustrative Trace (5)

An Illustrative Trace 5

Content Introduction Learning with Perfect Domain Theories: PROLOG-EBG Remarks on Explanation-Based Learning Discovering new features Summary

Remarks on Explanation-Based Learning Key properties: PROLOG-EBG produces justified general hypotheses by using prior knowledge to analyse individual examples The explanation about the way how an example satisfies the target concept determines which example attributes are relevant: the ones mentioned by the explanation Regressing the target concept to determine its weakest preimage with respect to the explanation allows deriving more general constraints on the values of relevant features Each learned Horn clause corresponds to a sufficient condition for satisfying the target concept The generality of the learned Horn clauses will depend on the formulation of the domain theory and on the sequence in which the training examples are considered Implicitly assumes that the domain theory is correct and complete

Remarks Explanation-Based Learning 2 Related perspectives to help to understand its capabilities and limitations: EBL as theory guided generalisation of examples: Rational generalisation from examples allows to avoid the bounds on sample complexity that occured in pure inductive learning EBL as example guided reformulation of theories: Method for reformulating the domain theory into more operational form: Creating rules that: Deductively follow the domain theory Classify the observed training examples in a single inference step

Remarks Explanation-Based Learning 3 Related perspectives to help to understand its capabilities and limitations: EBL is „just“ restating what the learner already „knows“: In what sense does this quality help to learn then? Knowledge reformulation: In many tasks the difference between what one knows in principle and what one can efficiently compute in practice may be great Situation: Complete perfect domain theory is already known to the (human) learner, and further learning is „simple“! So it's a matter of reformulating this knowledge into a form in which it can be used more effectively to select appropriate moves.

Remarks Explanation-Based Learning 4 Knowledge Compilation: EBL involves reformulating the domain theory to produce general rules that classify examples in a single inference step

Discovering new features Interesting capability: Ability to formulate new features that are not explicitly in the description of the training examples but that are needed to describe the general rule underlying the training examples This „feature“ is similarly represented by the hidden units of neural networks Like the BACKPROPAGATION algorithm, PROLOG_EBG automatically formulates such features in its attempt to fit the training data BUT: In neural networks it's developed in a statistical process PROLOG-EBG it's derived in an analytical process Example: derives the feature

Content Introduction Learning with Perfect Domain Theories: PROLOG-EBG Remarks on Explanation-Based Learning Summary

Summary PROLOG-EBG Uses first order Horn clauses in its domain theory and in its learned hypotheses The explanation is a PROLOG proof The hypothesis extracted from the explanation is the weakest preimage of this proof Analytical learning methods construct useful intermediate features as a side effect of analysing individual training examples. Other deductive learning procedures can extend the deductive closure of their domain. PRODIGY and SOAR have demonstrated the utility of explanation based learning methods for automatically acquiring effective search control knowledge that speeds up problem solving Disadvantage: purely deductive implementations such as PROLOG-EBG produce a correct output if the domain theory is also correct