Learning with recurrent neural networks [electronic resource] / Barbara Hammer
- Author:
- Hammer, Barbara, 1970-
- Published:
- London ; New York : Springer, [2000]
- Copyright Date:
- ©2000
- Physical Description:
- x, 148 pages : illustrations ; 24 cm.
Access Online
- Series:
- Restrictions on Access:
- License restrictions may limit access.
- Contents:
- Machine generated contents note: 1.Introduction -- 2.Recurrent and Folding Networks -- 2.1.Definitions -- 2.2.Training -- 2.3.Background -- 2.4.Applications -- 3.Approximation Ability -- 3.1.Foundations -- 3.2.Approximation in Probability -- 3.3.Approximation in the Maximum Norm -- 3.4.Discussion and Open Questions -- 4.Learnability -- 4.1.The Learning Scenario -- 4.2.PAC Learnability -- 4.3.Bounds on the VC-Dimension of Folding Networks -- 4.4.Consequences for Learnability -- 4.5.Lower Bounds for the LRAAM -- 4.6.Discussion and Open Questions -- 5.Complexity -- 5.1.The Loading Problem -- 5.2.The Perceptron Case -- 5.3.The Sigmoidal Case -- 5.4.Discussion and Open Questions -- 6.Conclusion.
- Subject(s):
- Genre(s):
- ISBN:
- 185233343X (pbk. : alk. paper)
- Bibliography Note:
- Includes bibliographical references (pages [137]-143) and index.
View MARC record | catkey: 7852624