Automatic classification of objects

---

ISBN 83-87674-83-4
Author: Katarzyna Stąpor
Publisher: EXIT

---

About the book

The book discusses the basic elements of automatic object classification systems. Within the supervised classification, empirical Bayes classifiers and directly defined classifiers are presented, including the supporting vector machine and the multi-layer perceptron. Among the grouping algorithms, selected algorithms of iterative optimization, density and graph algorithms as well as algorithms using neural networks are presented. All discussed algorithms are illustrated with practical examples with the participation of generated learning sets. In addition, applications of classification algorithms in computer vision systems are shown: recognition of geodetic maps and support of the diagnosis of glaucoma in ophthalmology.

The book will be useful to a wide group of readers involved in the design and creation of automatic object recognition systems in various fields, as well as students of the fields of Information Technology, Automation and Robotics and Bioengineering.

---

Table of Contents

1. Introduction

1.1. Examples of the classification task
1.2. Generalization, understatement and overfitting
1.3. Approaches to the classification of objects

2. Elements of the classification task

2.1. Object representation methods
2.2. The task of supervised classification
2.3. Task of unattended classification

3. Empirical Bayes classifiers

3.1. Bayes decision rule
3.2. Parametric classifiers
3.3. Non-parametric classifiers

4. Classifiers defined in a direct way

4.1. Linear classifiers
4.2. Generalized linear classifiers

5. Classifiers defined by symbolic structures

5.1. Basic symbolic structures
5.2. Structured matching of strings
5.3. Structural adjustment of graphs based on isomorphism
5.4. Structural adjustment of graphs based on homeomorphism
5.5. Structural adjustment based on knowledge representation methods

6. Classifiers defined by grammar

6.1. String grammars
6.2. Extensions of string grammar
6.3. The algorithm of Earley's syntactic analysis
6.4. Syntactic analysis with error correction
6.5. Syntactic analyzer of line drawings

7. Division grouping algorithms

7.1. Iterative optimization algorithms
7.2. Density algorithms
7.3. Graph algorithms
7.4. Algorithms using neural networks
7.5. Algorithms based on matching the statistical model
7.6. Grouping validation

8. Hierarchical grouping algorithms

8.1. Agglomeration matrix algorithms
8.2. Grouping of large data sets

9. Model selection and its evaluation

9.1. Methods using statistical theory of learning
9.2. Experimental methods

10. Extraction and selection of features

10.1. Initial data processing
10.2. Extraction of features
10.3. Selection of features
10.4. Selection of features using genetic algorithms