Keywords:
Relational databases.
;
Electronic books.
Description / Table of Contents:
What is knowledge and how is it represented? This introductory textbook presents relational methods in machine learning. Its clear and precise presentation is ideal for undergraduate computer science students and it will also interest those who study artificial intelligence or machine learning at the graduate level.
Type of Medium:
Online Resource
Pages:
1 online resource (280 pages)
Edition:
1st ed.
ISBN:
9781139514682
URL:
https://ebookcentral.proquest.com/lib/geomar/detail.action?docID=944687
DDC:
006.31
Language:
English
Note:
Cover -- Relational Knowledge Discovery -- Title -- Copyright -- Contents -- About this book -- What it is about -- How it is organised -- Thanks to: -- Chapter 1: Introduction -- 1.1 Motivation -- 1.1.1 Different kinds of learning -- 1.1.2 Applications -- 1.2 Related disciplines -- 1.2.1 Codes and compression -- 1.2.2 Information theory -- 1.2.3 Minimum description length -- 1.2.4 Kolmogorov complexity -- 1.2.5 Probability theory -- Conclusion -- Chapter 2: Relational knowledge -- 2.1 Objects and their attributes -- 2.1.1 Collections of things: sets -- 2.1.2 Properties of things: relations -- 2.1.3 Special properties of relations -- 2.1.4 Information systems -- 2.1.5 Structured sets -- 2.1.6 Probabilities -- 2.1.7 Relation algebra -- 2.2 Knowledge structures -- 2.2.1 Concepts, equivalence relations, and knowledge -- 2.2.2 Operations on equivalence relations -- 2.2.3 Indiscernability and knowledge -- Chapter 3: From data to hypotheses -- 3.1 Representation -- 3.2 Changing the representation -- 3.2.1 Linear separability -- 3.3 Samples -- 3.4 Evaluation of hypotheses -- 3.4.1 Error sets and error measures -- 3.4.2 Precision, accuracy, and others -- 3.5 Learning -- 3.6 Bias -- 3.7 Overfitting -- 3.8 Summary -- Chapter 4: Clustering -- 4.1 Concepts as sets of objects -- 4.2 k-nearest neighbours -- 4.3 k-means clustering -- 4.4 Incremental concept formation -- 4.5 Relational clustering -- Chapter 5: Information gain -- 5.1 Entropy -- 5.2 Information and information gain -- 5.2.1 Entropy -- 5.2.2 Information -- 5.3 Induction of decision trees -- 5.3.1 Hunt's classifier trees and Quinlan's ID3 -- 5.4 Gain again -- 5.5 Pruning -- 5.5.1 Reduced error pruning -- 5.5.2 Rule-based post-pruning -- 5.6 Conclusion -- Chapter 6: Rough set theory -- 6.1 Knowledge and discernability -- 6.2 Rough knowledge -- 6.2.1 Rough approximations -- 6.2.2 Degrees of roughness.
,
6.3 Rough knowledge structures -- 6.4 Relative knowledge -- 6.5 Knowledge discovery -- 6.5.1 Utility -- 6.5.2 Attribute significance -- 6.6 Conclusion -- Chapter 7: Inductive logic learning -- 7.1 From information systems to logic programs -- 7.1.1 Functions and relations -- 7.1.2 Semantics of first order logic -- 7.1.3 Deduction -- 7.2 Horn logic -- 7.2.1 Logic programs -- 7.2.2 Induction of logic programs -- 7.2.3 Entailment, generality, and subsumption -- 7.3 Heuristic rule induction -- 7.3.1 Refinement operators on H -- 7.3.2 Heuristic refinement -- 7.4 Inducing Horn theories from data -- 7.4.1 Syntactic generalisation revisited -- 7.4.2 Inverting resolution -- 7.4.3 Semantic biases -- 7.4.4 Inverted entailment -- 7.5 Summary -- Chapter 8: Learning and ensemble learning -- 8.1 Learnability -- 8.1.1 Probably approximately correct learning -- 8.1.2 Learnability and learning algorithms -- 8.2 Decomposing the learning problem -- 8.2.1 Bagging -- 8.3 Improving by focusing on errors -- 8.4 A relational view on ensemble learning -- 8.4.1 Dividing the sample set -- 8.4.2 Focusing on errors -- 8.5 Summary -- Chapter 9: The logic of knowledge -- 9.1 Knowledge representation -- 9.2 Learning -- 9.2.1 Clustering -- 9.2.2 Decision trees -- 9.2.3 Rough sets -- 9.2.4 Inductive logic programming -- 9.3 Summary -- Notation -- References -- Index.
