Fusion Methods for Unsupervised Learning Ensembles

• Publisher: Springer
• Year: 2011
• Tongue: English
• Leaves: 153
• Series: Studies in Computational Intelligence 322
• Category: Library

✦ Synopsis

As well as a selection of extended peer-reviewed papers from the 4th International IEEE Conference on Intelligent Systems, this volume is complemented by relevant work from leading figures who did not attend, and covers virtually all aspects of IS development.

✦ Table of Contents

Cover......Page 1
Front Matter......Page 2
WeVoS Algorithm......Page 3
Results Combination......Page 4
Discussion......Page 5
Application of WeVoS to Different Models......Page 6
Ensemble Models......Page 7
Quality Measures......Page 8
Comparison of Fusion Algorithms over a 1-D SOM......Page 9
Real Life Data Set: Food Industry Application......Page 12
Linear Combinations......Page 14
Contributions......Page 17
The Scale Invariant Map......Page 19
The Biological Neuron......Page 22
Unsupervised Learning......Page 25
Comparison between Single Model and Ensemble as Classifiers......Page 11
Comparison between Fusion by Distance and Fusion by Similarity Algorithms......Page 13
Ensembles of Artificial Neural Networks......Page 15
Background......Page 16
Organization......Page 18
Comparison of Fusion Algorithms over the 2-D SOM......Page 10
Comparison between Bagging and Boosting as Ensemble Training Algorithm......Page 21
Principal Component Analysis......Page 28
The Human Learning Process......Page 20
Learning Algorithms in Neural Networks......Page 24
Hebbian Learning and Statistics......Page 27
Artificial Neural Networks......Page 23
Hebbian Learning......Page 26
Negative Feedback Network......Page 31
The Self-Organizing Map......Page 33
The Visually Induced SOM......Page 36
The Scale Invariant Map......Page 38
Assessing Quality of Training of Topology Preserving Models......Page 41
Conclusions......Page 44
The Classification Problem......Page 45
Ensemble General Concepts......Page 46
Bagging......Page 48
Boosting......Page 51
Mixture of Experts......Page 55
Selection......Page 57
Linear Combinations......Page 58
Ensembles of Artificial Neural Networks......Page 59
Unsupervised ANNs......Page 60
Conclusions......Page 61
Introduction......Page 62
Ensemble Construction......Page 64
Results Combination......Page 65
Experiments and Results......Page 66
Artificial Data Set......Page 67
Real Life Data Set: Liver Disorder Data Set......Page 72
Real Life Data Set: Food Industry Application......Page 73
ANNs Approach......Page 78
Conclusions......Page 79
Introduction......Page 80
Topology-Preserving Map Combination Models......Page 81
Previously Proposed Models for SOM Ensemble Summarization......Page 82
Novel Proposed Model: Superposition......Page 87
Discussion of the Fusion Models......Page 89
Comparison between Single Model and Ensemble as Classifiers......Page 90
Comparison between Fusion by Distance and Fusion by Similarity Algorithms......Page 92
Comparison between Fusion by Distance and Superposition Algorithms......Page 96
Comparison between Bagging and Boosting as Ensemble Training Algorithm......Page 100
Food Industry Application......Page 104
Conclusions......Page 107
Introduction......Page 108
The Weighted Voting Superposition Algorithm......Page 109
WeVoS Algorithm......Page 110
Discussion......Page 112
Application of WeVoS to Different Models......Page 113
Ensemble Models......Page 114
Quality Measures......Page 115
Comparison of Fusion Algorithms over a 1-D SOM......Page 116
Comparison of Fusion Algorithms over the 2-D SOM......Page 117
Comparison of Fusion Algorithms over the ViSOM......Page 124
Comparison of Fusion Algorithms over the SIM and Max-SIM......Page 125
Comparison of Fusion Algorithms When Combined with Boosting......Page 127
Food Industry Application......Page 131
Conclusions......Page 134
Future Work......Page 135
Concluding Remarks......Page 136
Future Research Work......Page 137
Back Matter......Page 139