Amazing Books
Temporary Blank

Monday, October 10, 2011

Nature-inspired Methods in Chemometrics Genetic Algorithms and Artificial Neural Networks






CONTENTS
PREFACE vii
LIST OF CONTRIBUTORS xvii
PART I: GENETIC ALGORITHMS 1
CHAPTER 1 3
GENETIC ALGORITHMS AND BEYOND
(Brian T. Luke)
1 Introduction 3
2 Biological systems and the simple genetic algorithm (SGA) 5
3 Why do GAs work? 6
4 Creating a genetic algorithm 7
4.1 Determining a fitness function 7
4.2 The genetic vector 8
4.3 Creating an initial population 13
4.4 Selection schemes 14
4.5 Mating operators 16
4.6 Mutation operators 23
4.7 Maturation operators 25
4.8 Processing offspring 26
4.9 Termination metrics 27
5 Exploration versus exploitation 28
5.1 The genetic vector 29
5.2 The initial population 30
5.3 Selection schemes 31
5.4 Mating operators 33
5.5 Mutation operators 34
5.6 Maturation operators 34
5.7 Processing offspring 34
5.8 Balancing exploration and exploitation 36
6 Other population-based methods 40
6.1 Parallel GA 41
6.2 Adaptive parallel GA 41
6.3 Meta-GA 42
6.4 Messy GA 42
6.5 Delta coding GA 43
6.6 Tabu search and Gibbs sampling 43
6.7 Evolutionary programming 44
6.8 Evolution strategies 44
6.9 Ant colony optimization 45
6.10 Particle swarm optimization 46
7 Conclusions 48
CHAPTER 2 55
HYBRID GENETIC ALGORITHMS
(D. Brynn Hibbert)
1 Introduction 55
2 The approach to hybridization 55
2.1 Levels of interaction 56
2.2 A simple classification 57
3 Why hybridize? 57
4 Detailed examples 59
4.1 Genetic algorithm with local optimizer 59
4.2 Genetic algorithm–artificial neural network hybrid
optimizing quantitative structure–activity relationships 62
4.3 Non-linear partial least squares regression with optimization
of the inner relation function by a genetic algorithm 63
4.4 The use of a clustering algorithm in a genetic algorithm 64
5 Conclusion 66
CHAPTER 3 69
ROBUST SOFT SENSOR DEVELOPMENT USING
GENETIC PROGRAMMING
(Arthur K. Kordon, Guido F. Smits, Alex N. Kalos,
and Elsa M. Jordaan)
1 Introduction 69
2 Soft sensors in industry 71
2.1 Assumptions for soft sensors development 72
2.2 Economic benefits from soft sensors 73
2.3 Soft sensor application areas 74
2.4 Soft sensor vendors 75
3 Requirements for robust soft sensors 76
3.1 Lessons from industrial applications 76
3.2 Design requirements for robust soft sensors 77
4 Selected approaches for effective soft sensors development 80
4.1 Stacked analytical neural networks 80
4.2 Support vector machines 85
5 Genetic programming in soft sensors development 90
5.1 The nature of genetic programming 90
5.2 Solving problems with genetic programming 96
5.3 Advantages of genetic programming in soft
sensors development and implementation 98
6 Integrated methodology 99
6.1 Variable selection by analytical neural networks 100
6.2 Data condensation by support vector machines 101
6.3 Inferential model generation by genetic programming 102
6.4 On-line implementation and model self-assessment 102
7 Soft sensor for emission estimation: a case study 103
8 Conclusions 105
CHAPTER 4 109
GENETIC ALGORITHMS IN MOLECULAR MODELLING:
A REVIEW
(Alessandro Maiocchi)
1 Introduction 109
2 Molecular modelling and genetic algorithms 110
2.1 How to represent molecular structures and their conformations 111
3 Small and medium-sized molecule conformational search 114
4 Constrained conformational space searches 119
4.1 NMR-derived distance constraints 120
4.2 Pharmacophore-derived constraints 121
4.3 Constrained conformational search by chemical feature superposition 122
5 The protein-ligand docking problem 124
5.1 The scoring functions 126
5.2 Protein–ligand docking with genetic algorithms 127
6 Protein structure prediction with genetic algorithms 131
7 Conclusions 134
CHAPTER 5 141
MOBYDIGS: SOFTWARE FOR REGRESSION AND
CLASSIFICATION MODELS BY GENETIC ALGORITHMS
(Roberto Todeschini, Viviana Consonni, Andrea Mauri
and Manuela Pavan)
1 Introduction 141
2 Population definition 143
3 Tabu list 143
4 Random variables 144
5 Parent selection 145
6 Crossover/mutation trade-off 145
7 Selection pressure and crossover/mutation trade-off influence 148
8 RQK fitness functions 151
9 Evolution of the populations 154
10 Model distance 155
11 The software MobyDigs 158
11.1 The data setup 15811.2 GA setup 159
11.3 Population evolution view 161
11.4 Modify a single population evolution 162
11.5 Modify multiple population evolution 163
11.6 Analysis of the final models 164
11.7 Variable frequency analysis 165
11.8 Saving results 166
CHAPTER 6 169
GENETIC ALGORITHM-PLS AS A TOOL FOR
WAVELENGTH SELECTION IN SPECTRAL DATA SETS
(Riccardo Leardi)
1 Introduction 169
2 The problem of variable selection 170
3 GA applied to variable selection 172
3.1 Initiation of population 172
3.2 Reproduction and mutation 173
3.3 Insertion of new chromosomes 173
3.4 Control of replicates 174
3.5 Influence of the different parameters 174
3.6 Check of subsets 175
3.7 Hybridisation with stepwise selection 176
4 Evolution of the genetic algorithm 176
4.1 The application of randomisation tests 176
4.2 The optimisation of a GA run 177
4.3 Why a single run is not enough 177
4.4 How to take into account the autocorrelation
among the spectral variables 178
5 Pretreatment and scaling 181
6 Maximum number of variables 182
7 Examples 183
7.1 Data set Soy 183
7.2 Data set Additives 190
8 Conclusions 194
PART II: ARTIFICIAL NEURAL NETWORKS 197
CHAPTER 7 199
BASICS OF ARTIFICIAL NEURAL NETWORKS
(Jure Zupan)
1 Introduction 199
2 Basic concepts 200
Contents xii2.1 Neuron 200
2.2 Network of neurons 202
3 Error backpropagation ANNs 204
4 Kohonen ANNs 206
4.1 Basic design 206
4.2 Self-organized maps (SOMs) 210
5 Counterpropagation ANNs 213
6 Radial basis function (RBF) networks 216
7 Learning by ANNs 220
8 Applications 223
8.1 Classification 223
8.2 Mapping 224
8.3 Modeling 225
9 Conclusions 226
CHAPTER 8 231
ARTIFICIAL NEURAL NETWORKS IN MOLECULAR
STRUCTURES—PROPERTY STUDIES
(Marjana Novic and Marjan Vracko)
1 Introduction 231
2 Molecular descriptors 231
3 Counter propagation neural network 233
3.1 Architecture of a counter propagation neural network 233
3.2 Learning in the Kohonen and output layers 235
3.3 Counter propagation neural network as a tool in QSAR 236
4 Application in toxicology and drug design 237
4.1 A study of aquatic toxicity for the fathead minnow 237
4.2 A study of aquatic toxicity toward Tetrahymena pyriformis
on a set of 225 phenols 239
4.3 Example of QSAR modeling with receptor dependent descriptors 242
5 Conclusions 252
CHAPTER 9 257
NEURAL NETWORKS FOR THE CALIBRATION
OF VOLTAMMETRIC DATA
(Conrad Bessant and Edward Richards)
1 Introduction 257
2 Electroanalytical data 257
2.1 Amperometry 258
2.2 Pulsed amperometric detection 259
2.3 Voltammetry 259
2.4 Dual pulse staircase voltammetry 259
2.5 Representation of voltammetric data 261
Contents xiii3 Application of artificial neural networks to voltammetric data 261
3.1 Basic approach 262
3.2 Example of ANN calibration of voltammograms 263
3.3 Summary and conclusions 269
4 Genetic algorithms for optimisation of feed forward neural networks 269
4.1 Genes and chromosomes 269
4.2 Choosing parents for the next generation 270
4.3 Results of ANN optimisation by GA 272
4.4 Comparison of optimisation methods 277
5 Conclusions 278
CHAPTER 10 281
NEURAL NETWORKS AND GENETIC ALGORITHMS
APPLICATIONS IN NUCLEAR MAGNETIC
RESONANCE (NMR) SPECTROSCOPY
(Reinhard Meusinger and Uwe Himmelreich)
1 Introduction 281
2 NMR spectroscopy 283
3 Neural networks applications 285
3.1 Classification 286
3.2 Prediction of properties 290
4 Genetic algorithms 303
4.1 Data processing 304
4.2 Structure determination 305
4.3 Structure prediction 308
4.4 Classification 308
4.5 Feature reduction 309
5 Biomedical NMR spectroscopy 309
6 Conclusion 315
CHAPTER 11 323
A QSAR MODEL FOR PREDICTING THE ACUTE
TOXICITY OF PESTICIDES TO GAMMARIDS
(James Devillers)
1 Introduction 323
2 Materials and methods 324
2.1 Toxicity data 324
2.2 Molecular descriptors 324
2.3 Statistical analyses 329
3 Results and discussion 330
3.1 PLS model 330
3.2 ANN model 332
4 Conclusions 338
Contents xivCONCLUSION 341
CHAPTER 12 343
APPLYING GENETIC ALGORITHMS AND NEURAL
NETWORKS TO CHEMOMETRIC PROBLEMS
(Brian T. Luke)
1 Introduction 343
2 Structure of the genetic algorithm 345
3 Results for the genetic algorithms 350
4 Structure of the neural network 362
5 Results for the neural network 365
6 Conclusions 373
INDEX 377

Another Neural Network Books
Another Artificial Intelligence Books
Another Genetic Algorithm Books
Download

No comments:

Post a Comment

Related Posts with Thumbnails
There was an error in this gadget

Put Your Ads Here!
There was an error in this gadget