Gaining degrees of freedom in subsymbolic learning

Theoretical Computer Science

Volumn 255, Issue 1-2, 2001, Pages 295-321

  Apolloni, B ^a   Malchiodi, D ^a  

^a UNIVERSITY OF MILAN   (Italy)

Author keywords

Approximate learning; Computational learning; Nested concept classes; Neural networks; Sentry functions

Indexed keywords

APPROXIMATION THEORY; COMPUTATIONAL COMPLEXITY; NEURAL NETWORKS; STATISTICS;

COMPUTATIONAL LEARNING; NESTED CONCEPT CLASSES; SENTRY FUNCTIONS;

LEARNING SYSTEMS;

EID: 0034905723     PISSN: 03043975     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0304-3975(99)00289-3     Document Type: Article

References (32)

1. Hints and the VC dimension
2. Learning from noisy examples
3. Design of algorithms for neural networks
4. PAC Learning of concept classes through the boundaries of their items
5. Approximation of optimization problems and learnability
6. Hybrid symbolic subsymbolic system for controlling a single link flexible arm
7. Learning fuzzy decision trees
8. Learning with a slowly changing distribution
9. What size net gives valid generalizations?
10. Learnability and the Vapnik-Chervonenkis dimension
11. A general lower-bound on the number of examples needed for learning
12. Unified integration of explicit knowledge and learning by example in recurrent networks
13. Tracking drifting concepts by minimizing disagreements
14. Efficient noise-tolerant learning from statistical queries
15. Learning in the presence of malicious errors
16. On learning boolean functions
17. Non parametric estimation II, Statistical equivalent blocks and tolerance regions - The continuous case
18. Non parametric estimation III, Statistical equivalent blocks and multivariate tolerance regions - The discontinuous case
19. A theory of the learnable
20. Some special Vapnik-Chervonenkis classes