Learning good prototypes for classification using filtering and abstraction of instances

Pattern Recognition

Volumn 35, Issue 7, 2002, Pages 1491-1506

  Lam, Wai ^a   Keung, Chi Kin ^a   Ling, Charles X ^b  

^a CHINESE UNIVERSITY OF HONG KONG   (Hong Kong)

^b UNIVERSITY OF WESTERN ONTARIO   (Canada)

Author keywords

Instance abstraction; Instance based learning; Machine learning; Prototype generation

Indexed keywords

ABSTRACTING; CLASSIFICATION (OF INFORMATION); INTEGRATION;

PROTOTYPES;

PATTERN RECOGNITION;

IMAGE CLASSIFICATION; IMAGE PROCESSING;

EID: 0036643351     PISSN: 00313203     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0031-3203(01)00131-5     Document Type: Article

References (27)

1. Multiple-instance learning for natural scene classification
2. Refinements to nearest-neighbor searching in k-dimensional trees
3. Prototype generation based on instance filtering and averaging
4. The condensed nearest neighbor rule
5. The reduced nearest neighbor rule
6. Asymptotic properties of nearest neighbor rules using edited data
7. Noise-tolerant instance-based learning algorithms
8. Instance-based learning algorithms
9. Selecting typical instances in instance-based learning
10. Instance pruning techniques
11. Reduction techniques for instance-based learning algorithm
12. An integrated instance-based learning algorithm
13. Finding prototypes for nearest neighbor classifiers
14. Learning about speech sounds: The Nexus project
15. Comparing instance-averaging with instance-filtering learning algorithms
16. A nearest hyperrectangle learning method
17. A hybrid nearest-neighbor and nearest-hyperrectangle algorithm
18. Unifying instance-based and ruled-based induction
19. Learning prototypical concept description
20. Symbolic nearest mean classifier
21. A weighted nearest neighbor algorithm for learning with symbolic feature
22. Multiple-prototype classifier design
23. Instance-family abstraction in memory-based language learning
24. Squashing flat files flatter
25. Toward memory-based reasoning