Predictive Modular Neural Networks

1. Front Matter

   Pages i-xi

   PDF

2. Introduction

   1. Introduction

      • Vassilios Petridis, Athanasios Kehagias

      Pages 1-7

3. Known Sources

   1. Front Matter

      Pages 9-9

      PDF

   2. Premonn Classification and Prediction

      • Vassilios Petridis, Athanasios Kehagias

      Pages 11-38

   3. Generalizations of the Basic Premonn

      • Vassilios Petridis, Athanasios Kehagias

      Pages 39-57

   4. Mathematical Analysis

      • Vassilios Petridis, Athanasios Kehagias

      Pages 59-80

   5. System Identification by the Predictive Modular Approach

      • Vassilios Petridis, Athanasios Kehagias

      Pages 81-97

4. Applications

   1. Front Matter

      Pages 99-99

      PDF

   2. Implementation Issues

      • Vassilios Petridis, Athanasios Kehagias

      Pages 101-107

   3. Classification of Visually Evoked Responses

      • Vassilios Petridis, Athanasios Kehagias

      Pages 109-122

   4. Prediction of Short Term Electric Loads

      • Vassilios Petridis, Athanasios Kehagias

      Pages 123-133

   5. Parameter Estimation for and Activated Sludge Process

      • Vassilios Petridis, Athanasios Kehagias

      Pages 135-145

5. Unknown Sources

   1. Front Matter

      Pages 147-147

      PDF

   2. Source Identification Algorithms

      • Vassilios Petridis, Athanasios Kehagias

      Pages 149-172

   3. Convergence of Parallel Data Allocation

      • Vassilios Petridis, Athanasios Kehagias

      Pages 173-207

   4. Convergence of Serial Data Allocation

      • Vassilios Petridis, Athanasios Kehagias

      Pages 209-245

6. Connections

   1. Front Matter

      Pages 247-247

      PDF

   2. Bibliographic Remarks

      • Vassilios Petridis, Athanasios Kehagias

      Pages 249-266

   3. Epilogue

      • Vassilios Petridis, Athanasios Kehagias

      Pages 267-269

7. Back Matter

   Pages 271-314

   PDF

The subject of this book is predictive modular neural networks and their ap­ plication to time series problems: classification, prediction and identification. The intended audience is researchers and graduate students in the fields of neural networks, computer science, statistical pattern recognition, statistics, control theory and econometrics. Biologists, neurophysiologists and medical engineers may also find this book interesting. In the last decade the neural networks community has shown intense interest in both modular methods and time series problems. Similar interest has been expressed for many years in other fields as well, most notably in statistics, control theory, econometrics etc. There is a considerable overlap (not always recognized) of ideas and methods between these fields. Modular neural networks come by many other names, for instance multiple models, local models and mixtures of experts. The basic idea is to independently develop several "subnetworks" (modules), which may perform the same or re­ lated tasks, and then use an "appropriate" method for combining the outputs of the subnetworks. Some of the expected advantages of this approach (when compared with the use of "lumped" or "monolithic" networks) are: superior performance, reduced development time and greater flexibility. For instance, if a module is removed from the network and replaced by a new module (which may perform the same task more efficiently), it should not be necessary to retrain the aggregate network.

• algorithms
• artificial intelligence
• classification
• computer
• control engineering
• convergence
• data mining
• design
• network
• neural networks
• time series analysis

• Aristotle University of Thessaloniki, Greece

  Vassilios Petridis

• American College of Thessaloniki and Aristotle University of Thessaloniki, Greece

  Athanasios Kehagias

Policies and ethics