Java Object Oriented Neural Engine. The Complete Guide

Marrone P.

Table of contents :
Cover……Page 1
TOC……Page 2
Foreword……Page 5
1.2.1 Custom systems……Page 6
1.3 About this Guide……Page 7
1.4 Acknowledgements……Page 9
2.2.1 The Core Engine……Page 10
2.2.2 The GUI Editor……Page 12
2.3.1 Prerequisites……Page 18
2.3.3 Getting the CVS sources……Page 19
2.3.4 Compiling……Page 20
3.2 The Transport Mechanism……Page 21
3.3.1.1 The Linear Layer……Page 24
3.3.1.5 The Delay Layer……Page 25
3.3.1.6 The Context Layer……Page 26
3.3.1.8 The Gaussian Layer……Page 27
3.3.2.1 The Direct Synapse……Page 29
3.3.2.3 The Delayed Synapse……Page 30
3.3.2.4 The Kohonen Synapse……Page 31
3.4 The Monitor: a central point to control the neural network……Page 32
3.4.1 The Monitor as a container of the NN Parameters……Page 33
3.4.2 The Monitor as the NN controller……Page 34
3.4.3 Managing the events……Page 35
3.4.4.1 The Pattern……Page 37
3.5 Technical details……Page 38
3.5.1 The Layer abstract class……Page 39
3.5.1.1 The Recall Phase……Page 40
3.5.2 Connecting a Synapse to a Layer……Page 41
3.5.3 The Synapse abstract class……Page 42
4.1 The Input mechanism……Page 45
4.2 The Output: using the outcome of a neural network……Page 46
4.3 The Switching Mechanism……Page 47
4.4 The Validation Mechanism……Page 48
4.5 Technical details……Page 49
4.5.1 The StreamInputSynapse……Page 51
4.5.3 The Switching mechanism’s object model……Page 53
4.5.3.2 The OutputSwitchSynapse……Page 54
4.5.3.3 The LearningSwitch……Page 55
5.1 The Teacher component……Page 56
5.2 Technical details……Page 58
6.2 The Monitor Plugins……Page 59
6.3 The Scripting Mechanism……Page 61
6.4.1 The Input Plugins object model……Page 62
6.4.2 The Monitor Plugin object model……Page 65
6.4.3 The Scripting mechanism object model……Page 67
7.1 The NeuralNet object……Page 70
7.2 Technical details……Page 72
8.1.1.1 Preprocessing……Page 73
8.2.1.1 Example: a character recognition system……Page 76
9.1 A simple (but useless) neural network……Page 83
9.2 A real implementation: the XOR problem…….Page 84
9.3.1 The simplest way……Page 88
9.3.2 Using a NeuralNet object……Page 89
9.4.1 Writing the results to an output file……Page 92
9.4.3 Using multiple input patterns……Page 93
9.4.4 Using only one input pattern……Page 96
10 The LGPL Licence……Page 98