Learning Classifier Systems: International Workshops, IWLCS 2003-2005, Revised Selected PapersTim Kovacs, Xavier Llorà, Keiki Takadama, Pier Luca Lanzi, Wolfgang Stolzmann, Stewart W. Wilson The work embodied in this volume was presented across three consecutive e- tions of the International Workshop on Learning Classi?er Systems that took place in Chicago (2003), Seattle (2004), and Washington (2005). The Genetic and Evolutionary Computation Conference, the main ACM SIGEvo conference, hosted these three editions. The topics presented in this volume summarize the wide spectrum of interests of the Learning Classi?er Systems (LCS) community. The topics range from theoretical analysis of mechanisms to practical cons- eration for successful application of such techniques to everyday data-mining tasks. When we started editing this volume, we faced the choice of organizing the contents in a purely chronologicalfashion or as a sequence of related topics that help walk the reader across the di?erent areas. In the end we decided to or- nize the contents by area, breaking the time-line a little. This is not a simple endeavor as we can organize the material using multiple criteria. The tax- omy below is our humble e?ort to provide a coherent grouping. Needless to say, some works may fall in more than one category. The four areas are as follows: Knowledge representation. These chapters elaborate on the knowledge r- resentations used in LCS. Knowledge representation is a key issue in any learning system and has implications for what it is possible to learn and what mechanisms shouldbe used. Four chapters analyze di?erent knowledge representations and the LCS methods used to manipulate them. |
Contents
| 1 | |
Use of Learning Classifier Systemfor Inferring Natural Language Grammar | 17 |
Backpropagation in AccuracyBased Neural LearningClassifier Systems | 25 |
A Study Using the Compact Classifier System | 40 |
Bloat Control and Generalization Pressure Usingthe Minimum Description Length Principle for aPittsburgh Approach Learning Classifier System | 59 |
Postprocessing Clustering to DecreaseVariability in XCS Induced Rulesets | 80 |
Learning Classifier System Ensemble forIncremental Medical Instances | 93 |
Effect of Pure ErrorBased Fitness in XCS | 104 |
Three Methods forCovering Missing Input Data in XCS | 181 |
Learning to Create Novel ProblemSolvingAlgorithms Constructed from SimplerAlgorithmic Ingredients | 193 |
Adaptive Value Function Approximations inClassifier Systems | 219 |
Three Architectures for Continuous Action | 239 |
A Formal Relationship Between Ant Colony Optimizersand Classifier Systems | 258 |
A Sensitivity Analysis | 270 |
Comparing XCS with GAssist | 282 |
Improving the Performance of a PittsburghLearning Classifier System Using a Default Rule | 291 |
A Fuzzy System to Control Exploration Rate in XCS | 115 |
Counter Example for QBucketBrigadeUnder Prediction Problem | 128 |
An Experimental Comparison BetweenATNoSFERES and ACS | 144 |
A Preliminary Study | 161 |
Using XCS to Describe ContinuousValuedProblem Spaces | 308 |
A Tool for Experimentation and Visualization | 333 |
| 345 | |
Common terms and phrases
accuracy accurate action adaptive addition algorithm analysis applied approach approximation attributes average better clusters coding combination compared complexity Computation condition Conference covering cycles dataset default defined described distance distribution domain effect encoding environment error evaluate evolved example experiments exploration Figure Finally fitness follows function Genetic Algorithms given heuristics imbalance important improve increase individual initial input instances interval iterations Lanzi Learning Classifier Systems majority match measure method minority class missing node obtained operator optimal parameters payoff performance points population positive possible prediction presented probability problem Proceedings proposed random range regions reinforcement represent representation respect reward rules runs sample scheme selection shown shows similar simple single solution space specific step Table tile University update variable weight Wilson