Sensitivity Analysis for Neural Networks

1. Front Matter

   Pages i-viii

   PDF

2. Introduction to Neural Networks

   • Daniel S. Yeung, Ian Cloete, Daming Shi, Wing W.Y. Ng

   Pages 1-15

3. Principles of Sensitivity Analysis

   • Daniel S. Yeung, Ian Cloete, Daming Shi, Wing W.Y. Ng

   Pages 17-24

4. Hyper-Rectangle Model

   • Daniel S. Yeung, Ian Cloete, Daming Shi, Wing W.Y. Ng

   Pages 25-27

5. Sensitivity Analysis with Parameterized Activation Function

   • Daniel S. Yeung, Ian Cloete, Daming Shi, Wing W.Y. Ng

   Pages 29-31

6. Localized Generalization Error Model

   • Daniel S. Yeung, Ian Cloete, Daming Shi, Wing W.Y. Ng

   Pages 33-46

7. Critical Vector Learning for RBF Networks

   • Daniel S. Yeung, Ian Cloete, Daming Shi, Wing W.Y. Ng

   Pages 47-53

8. Sensitivity Analysis of Prior Knowledge1

   • Daniel S. Yeung, Ian Cloete, Daming Shi, Wing W.Y. Ng

   Pages 55-67

9. Applications

   • Daniel S. Yeung, Ian Cloete, Daming Shi, Wing W.Y. Ng

   Pages 69-82

10. Back Matter

    Pages 83-86

    PDF

Artificial neural networks are used to model systems that receive inputs and produce outputs. The relationships between the inputs and outputs and the representation parameters are critical issues in the design of related engineering systems, and sensitivity analysis concerns methods for analyzing these relationships. Perturbations of neural networks are caused by machine imprecision, and they can be simulated by embedding disturbances in the original inputs or connection weights, allowing us to study the characteristics of a function under small perturbations of its parameters.

This is the first book to present a systematic description of sensitivity analysis methods for artificial neural networks. It covers sensitivity analysis of multilayer perceptron neural networks and radial basis function neural networks, two widely used models in the machine learning field. The authors examine the applications of such analysis in tasks such as feature selection, sample reduction, and network optimization. The book will be useful for engineers applying neural network sensitivity analysis to solve practical problems, and for researchers interested in foundational problems in neural networks.

• Adaline
• Backpropagation algorithm
• Hyperrectangle model
• Learning
• Multilayer perceptron (MLP)
• Neural networks
• Perceptron
• Perturbations
• Sensitivity analysis
• Supervised learning
• Unsupervised learning
• Vector learni
• machine learning
• optimization
• perception

From the reviews:

“Neural Networks are seen as an information paradigm inspired by the way the human brain processes information. … The book may be used by researchers in diverse domains, such as neural networks, machine learning, computer engineering, etc., facing problems connected to sensitivity analysis of neural networks.” (Florin Gorunescu, Zentralblatt MATH, Vol. 1189, 2010)

• School of Computer Science &, South China University of Technology, Guangzhou, China, People's Republic

  Daniel S. Yeung, Wing W. Y. Ng

• President's Office, International University in Germany, Bruchsal, Germany

  Ian Cloete

• Dept. Electrical Engineering, Kyungpook National University, Daegu, Korea, Republic of (South Korea)

  Daming Shi

Policies and ethics