Proceedings of the Fourth Annual Workshop on Computational Learning Theory, University of California, Santa Cruz, August 5-7, 1991
COLT
| Autor principal: | |
|---|---|
| Autores Corporativos: | , , |
| Otros Autores: | , , , , |
| Formato: | Libro electrónico |
| Idioma: | Inglés |
| Publicado: |
San Mateo, California :
Morgan Kaufmann Publishers, Inc
1991.
|
| Colección: | ACM Conferences
|
| Materias: | |
| Ver en Biblioteca Universitat Ramon Llull: | https://discovery.url.edu/permalink/34CSUC_URL/1im36ta/alma991009713680506719 |
Tabla de Contenidos:
- Front Cover; Computational Learning Theory; Copyright Page; Table of Contens; Foreword; Invited Talks; Learning and Generalization; Chapter 1. The Role of Learning in Autonomous Robots; Abstract; 1 INTRODUCTION; 2 AUTONOMOUS ROBOTS; 3 LEARNING; 4 EXAMPLES; 5 CONCLUSION; Acknowledgements; References; Session I; Chapter 2. Tracking Drifting Concepts Using Random Examples; Abstract; 1 Introduction; 2 Notation and Mathematical Preliminaries; 3 The Exception Tracking Strategy; 4 Upper bounds on tolerable amount of drift; 5 Increasingly unreliable evidence and hypothesis evaluation; 6 Conclusions
- 7 AcknowledgementsReferences; Chapter 3. Investigating the Distribution Assumptions in the Pac Learning Model; Abstract; 1 INTRODUCTION; 2 LEARNING WITH REASONABLE DISTRIBUTIONS; 3 LEARNING WHEN THE DISTRIBUTION OF EXAMPLES CHANGES; 4 CONCLUSIONS; Acknowledgement; References; Chapter 4. Simultaneous Learning of Concepts and Simultaneous Estimation of Probabilities; Abstract; 1 INTRODUCTION; 2 SIMULTANEOUS LEARNING; 3 SIMULTANEOUS ESTIMATION; 4 EMPIRICAL COVERS; 5 RELATING SIMULTANEOUS LEARNING AND SIMULTANEOUS ESTIMATION; 6 RELATING SIMULTANEOUS LEARNING AND ESTIMATION TO EMPIRICAL COVERINGS
- 7 A SUFFICIENT CONDITION FOR SIMULTANEOUS LEARNING8 THE DISTRIBUTION-FREE CASE; 9 CONCLUSION AND OPEN PROBLEMS; Acknowledgements; APPENDIX; REFERENCES; Chapter 5. Learning by Smoothing: a morphological approach; Abstract; 1 INTRODUCTION; 2 FUNDAMENTALS; 3 PROPERTIES; 4 COMMENTS; Acknowledgements; References; Session II; Chapter 6. Bounds on the Sample Complexity of Bayesian Learning Using Information Theory and the VC Dimension; Abstract; 1 Introduction; 2 Summary of results; 3 Notational conventions; 4 Instantaneous information gain and mistake probabilities
- 5 Bounding the mistake probabilities by the information gain6 Bounding the cumulative mistakes by the partition entropy; 7 Handling incorrect priors; 8 The average instantaneous information gain is decreasing; 9 Bayesian learning and the VC dimension: correct priors; 10 Bayesian learning and the VC dimension: incorrect priors; 11 Conclusions and future research; Acknowledgements; References; Chapter 7. Calculation of the Learning Curve of Bayes Optimal Classification Algorithm for Learning a Perceptron With Noise; Abstract; 1 Introduction; 2 Results; 3 Conclusion; Acknowledgements; References
- Chapter 8. Probably Almost Bayes DecisionsAbstract; 1 INTRODUCTION; 2 BAYES AND PROBABLY ALMOST BAYES DECISIONS; 3 PAC-ESTIMABLE DISTRIBUTION CLASSES; 4 INDEPENDENT BOOLEAN FEATURES; 5 DEPENDENT BOOLEAN FEATURES; 6 Conclusions; References; Session IV; Chapter 9. A Geometric Approach to Threshold Circuit Complexity; Abstract; 1 Introduction; 2 Uniqueness; 3 Generalized Spectrum Techniques; 4 On The Method of Correlations; 5 Miscellaneous; 6 Concluding Remarks; Acknowledgments; Appendix; References; Chapter 10. Learning Curves in Large Neural Networks; Abstract; 1 Introduction; 2 General Theory
- 3 The Annealed Approximation
