Neural Networks: Tricks of the Trade

Second Edition

ISBN: 978-3-642-35288-1 (Print) 978-3-642-35289-8 (Online)

Table of contents (39 chapters)

previous Page of 2
  1. Front Matter

    Pages -

  2. Introduction

    1. No Access

      Book Chapter

      Pages 1-5

      Introduction

  3. Speeding Learning

    1. No Access

      Book Chapter

      Pages 7-8

      Speeding Learning

    2. No Access

      Book Chapter

      Pages 9-48

      Efficient BackProp

  4. Regularization Techniques to Improve Generalization

    1. No Access

      Book Chapter

      Pages 49-51

      Regularization Techniques to Improve Generalization

    2. No Access

      Book Chapter

      Pages 53-67

      Early Stopping — But When?

    3. No Access

      Book Chapter

      Pages 69-89

      A Simple Trick for Estimating the Weight Decay Parameter

    4. No Access

      Book Chapter

      Pages 91-110

      Controlling the Hyperparameter Search in MacKay’s Bayesian Neural Network Framework

    5. No Access

      Book Chapter

      Pages 111-130

      Adaptive Regularization in Neural Network Modeling

    6. No Access

      Book Chapter

      Pages 131-137

      Large Ensemble Averaging

  5. Improving Network Models and Algorithmic Tricks

    1. No Access

      Book Chapter

      Pages 139-141

      Improving Network Models and Algorithmic Tricks

    2. No Access

      Book Chapter

      Pages 143-161

      Square Unit Augmented, Radially Extended, Multilayer Perceptrons

    3. No Access

      Book Chapter

      Pages 163-189

      A Dozen Tricks with Multitask Learning

    4. No Access

      Book Chapter

      Pages 191-203

      Solving the Ill-Conditioning in Neural Network Learning

    5. No Access

      Book Chapter

      Pages 205-223

      Centering Neural Network Gradient Factors

    6. No Access

      Book Chapter

      Pages 225-230

      Avoiding Roundoff Error in Backpropagating Derivatives

  6. Representing and Incorporating Prior Knowledge in Neural Network Training

    1. No Access

      Book Chapter

      Pages 231-233

      Representing and Incorporating Prior Knowledge in Neural Network Training

    2. No Access

      Book Chapter

      Pages 235-269

      Transformation Invariance in Pattern Recognition – Tangent Distance and Tangent Propagation

    3. No Access

      Book Chapter

      Pages 271-293

      Combining Neural Networks and Context-Driven Search for On-line, Printed Handwriting Recognition in the Newton

    4. No Access

      Book Chapter

      Pages 295-309

      Neural Network Classification and Prior Class Probabilities

    5. No Access

      Book Chapter

      Pages 311-338

      Applying Divide and Conquer to Large Scale Pattern Recognition Tasks

previous Page of 2