Applications of Evolutionary Computation

Name: Applications of Evolutionary Computation | EvoApplications 2011: EvoCOMPLEX, EvoGAMES, EvoIASP, EvoINTELLIGENCE, EvoNUM, and EvoSTOC, Torino, Italy, April 27-29, 2011, Proceedings, Part I
Brand: Springer
Price: 53.49 EUR
Availability: OnlineOnly

EvoApplications 2011: EvoCOMPLEX, EvoGAMES, EvoIASP, EvoINTELLIGENCE, EvoNUM, and EvoSTOC, Torino, Italy, April 27-29, 2011, Proceedings, Part I

Cecilia Di Chio Stefano Cagnoni Carlos Cotta Marc Ebner Aniko Ekart Anna I. Esparcia-Alcázar Juan J. Merelo Ferrante Neri Mike Preuss Hendrik Richter Julian Togelius Georgios N. Yannakakis(Herausgeber*in)

Springer (Verlag)

Erschienen am 27. April 2011

XXXI, 367 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-3-642-20525-5 (ISBN)

53,49 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Wasserzeichen-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

Weitere Details

Weitere Ausgaben

Inhalt

Title
Preface
Organization
Table of Contents
EvoCOMPLEX Contributions
Evolving L-Systems as an Intelligent Design Approach to Find Classes of Difficult-to-Solve Traveling Salesman Problem Instances
Introduction
Related Work
Problem Statement
Method and Algorithm
Instance Generator (L-Systems)
Fitness Function for an L-System
Modifying L-Systems
Algorithm Overview
Simulation Results
Discussion and Analysis
Conclusions
References
A Design Framework for Ultra-Large-Scale Autonomic Systems
Introduction
Related Work
Modeling ULS Autonomic Systems
Implementation of the Adaptive Evolutionary Framework for ULS Systems
Simulation Experiments
Conclusion
References
Stochastic Local Search to Automatically Design Boolean Networks with Maximally Distant Attractors
Introduction
Overview of Boolean Networks
Objective and Motivations
Related Work
Formulation of the Design Problem
Experimental Analysis
Conclusions and Future Work
References
A Study on the Mutation Rates of a Genetic Algorithm Interacting with a Sandpile
Introduction
State-of-the-Art
The Sandpile and the Mutation Operator
Test Set and Results
Conclusions and Future Work
References
On the Design of Boolean Network Robots
Introduction
Boolean Networks
Boolean Network Robotics
Design Methodology
Test Case: Phototaxis and Antiphototaxis
BN Setup
Design of BN Robot
Conclusion and Future Work
References
EvoGAMES Contributions
Multiple Tree for Partially Observable Monte-Carlo Tree Search
Introduction
The EXP3 Algorithm
Our Algorithm: Multiple Monte-Carlo Tree Search with EXP3 as a Bandit Tool
Experimental Results
Conclusion
References
Evolving Interesting Maps for a First Person Shooter
Introduction
Cube 2
Evolving Maps for an FPS Game
Fitness Function
Map Representations
Experiments
Conclusions and Future Work
References
Revisiting Monte-Carlo Tree Search on a Normal Form Game: NoGo
Introduction
A Brief Overview of Monte-Carlo Tree Search (MCTS)
Improvements over the Initial MCTS Algorithm
Rapid Action Value Estimates
Rate of Node Creation
Efficiency of the Playouts and Decisive Moves
Upper Confidence Trees
A Generic Monte-Carlo Algorithm for Normal Form Games
Scalability Analysis
Conclusion
References
Nested Look-Ahead Evolutionary Algorithm Based Planning for a Believable Diplomacy Bot
Introduction
Diplomacy Primer
Diplomacy Complexity
Total Positions on the Map
Unit Moves
Evolutionary Algorithm-Based Move Planning
Planning vs. Non-planning Stragotiator
Summary and Conclusions
References
Towards Procedural Strategy Game Generation: Evolving Complementary Unit Types
Introduction
The Strategy Game Description Language
Overall Structure
Evolving Complementary Unit Types
Method
Results
Analysis of Evolved Unit Type Sets
Discussion
References
Improving and Scaling Evolutionary Approaches to the MasterMind Problem
Introduction
Background
Introducing endgames
Experimental Results
Conclusions and Future Work
References
Training Neural Networks to Play Backgammon Variants Using Reinforcement Learning
Introduction
Background
TD($\lambda$) Algorithm
Neural Network Function Approximation
The Games Plakoto and Fevga
Plakoto
Fevga
Implementation
Learning Architecture
Determining the Learning Parameters
Empirical Evaluation
Experiments in Fevga
Experiments in Plakoto
Conclusion
References
Evolving Behaviour Trees for the Mario AI Competition Using Grammatical Evolution
Introduction
Relevant Literature
Grammatical Evolution
Example Mapping Process
The Mario AI Benchmark
Environment Information
Mario Effectors
Behaviour Trees
Introduction
Behaviour Trees for Mario
Incorporation into GE
Experiments
Setup
Results
Conclusions
References
Learning Chasing Behaviours of Non-Player Characters in Games Using SARSA
Background
Learning Chasing and Evading Behaviours Using SARSA
Problem Formulation
Experimental Design
Designing the Environment and Agents' Behaviours
Designing Reward and Penalty Signals
Evaluation Criteria
Results and Discussion
Discussion on the Approach and Design Choice
Conclusion
References
Driving Faster Than a Human Player
Introduction
Learning Target Speeds
The Possible Speed is a Function of Curvature
Experimental Results of the Offline Learning Process
Online Adaption of the Learned Behaviour
Experimental Comparison
How Does This Compare to Learning on the Track We Are Racing on?
Summary and Conclusions
References
Upper Confidence Trees with Short Term Partial Information
Introduction
Frameworks
Bounded Horizon Hidden Information Games (BHHIG)
Games with Simultaneous Actions (GSA)
Representing BHHIG as GSA and GSA as BHHIG
Complexity of Games with Simultaneous Actions
Upper Confidence Trees for Games with Simultaneous Actions
The Upper Confidence Tree Algorithm
Adapting UCT to the GSA Acyclic Case
Experiments
The EXP3 Algorithm
Scalability
Games against Humans
Conclusion
References
EvoIASP Contributions
Segmentation of Ultrasound Breast Images: Optimization of Algorithm Parameters
Introduction
Segmentation of Breast Ultrasounds Images
Unspeckle Filtering
GrowCut Algorithm
Method
Preprocessing
Detail Segmentation
Parameter Selection
Experiments and Results
Conclusions
References
A Hybrid Particle Swarm Optimisation with Differential Evolution Approach to Image Segmentation
Introduction
Background
A Hybrid PSO-DE Method for Image Segmentation
Experimental Design
Results and Discussion
Conclusions
References
Transparent, Online Image Pattern Classification Using a Learning Classifier System
Introduction
Classification with LCS
Approaches to Image Pattern Classification
Naïve Pixel-Based Conditions
Haar-Like Feature Conditions
Experimental Results
Datasets
Experimental Setup
Results
Discussion and Future Work
Conclusions
References
Tracking Multiple Targets with Adaptive Swarm Optimization
Introduction
The Key Subproblems of Target Tracking
Objective Function of Target Identification
The Mathematical Model of Tracking the Target
Tracking Two Conflicting Targets in the Adjacent Area
Tracking Multiple Targets by Swarm Optimization
Experiments
Conclusions
References
Advanced Metaheuristic Approaches and Population Doping for a Novel Modeling-Based Method of Positron Emission Tomography Data Analysis
Introduction
Problem Description
Metaheutistic Approaches
Differential Evolution with Self-Adaptive Control Parameters
Multiple Trajectory Search
Comprehensive Learning Particle Swarm Optimizer
Intelligent Sampling for Metaheuristics
Numerical Results
Conclusion
References
Evolutionary Synthesis of a Trajectory Integrator for an Analogue Brain-Computer Interface Mouse
Introduction
BCI Mouse
GP System and Parameter Settings
Experimental Results
Conclusions
References
Automatic Selection of Pareto-Optimal Topologies of Hidden Markov Models Using Multicriteria Evolutionary Algorithms
Introduction
Background
Hidden Markov Models
Multi-Objective Optimization and Pareto Optimality
Pareto Optimal HMMs (POHMM) Topologies
Experiments
Bird Species Corpus
PPPC Corpus
Results
Research Scenarios
Miscellaneous Evolutionary Issues
Discussion, Conclusions and Future Work
References
EvoINTELLIGENCE Contributions
When Novelty Is Not Enough
Introduction
Novelty Search
Experiments
The Deceptive Tartarus Problem
Setup
Results
Discussion
References
A Part-Of-Speech Lexicographic Encoding for an Evolutionary Word Sense Disambiguation Approach
Introduction
Problem Description
The Evolutionary Approach
Context Encoding
Experiments and Results
Conclusion and Future Work
References
Migrating Birds Optimization: A New Meta-heuristic Approach and Its Application to the Quadratic Assignment Problem
Introduction
Migration of Birds
The Migrating Birds Optimization Algorithm
Application
Summary, Conclusions and Future Work
References
EvoNUM Contributions
Opposition-Based Learning in Compact Differential Evolution
Introduction
Compact Opposition-Based Differential Evolution
Numerical Results
Conclusion
References
Data Mining Using Unguided Symbolic Regression on a Blast Furnace Dataset
Introduction
Variable Relevance Metrics for GP
Extension of Frequency-Based Variable Relevance for GP
Experiments
Blast Furnace Dataset
Algorithmic Settings
Results
Variable Interaction Network
Detailed Results
Conclusion
References
DISPAR-Tournament: A Parallel Population Reduction Operator That Behaves Like a Tournament
Introduction
Tournament Selection Variants
Tournament "With Replacement"
Tournament "Without Replacement"
Disjoint Set Parallel (DISPAR) Tournament
Comparing DISPAR-Tournament with Standard Tournament
Metrics and Setup of Experimental Comparisons
Effect of Population Size on Tournament
Implementation
Conclusion and Future Work
References
Global Characterization of the CEC 2005 Fitness Landscapes Using Fitness-Distance Analysis
Introduction
Landscape Topologies and Their Impact on Continuous Black-Box Search Heuristics
Fitness-Distance Analysis
Fitness-Distance Analysis for Continuous Black-Box Landscapes
Characterization of the CEC 2005 Benchmark Test Suite
Fitness-Distance Plots and Distance Distributions
The Fitness-Distance Correlation Coefficients
Conclusions and Future Work
References
A Framework for Multi-model EDAs with Model Recombination
Introduction
Related Work
The Framework
The Multi-model EDA
The Real-Valued Multi-Model EDA (RVMMEDA)
A Framework Unifying EAs and EDAs
Experiments and Results
Experimental Settings
Experimental Results
Conclusions and Future Work
References
EvoSTOC Contributions
An Investigation of Selection Hyper-heuristics in Dynamic Environments
Introduction
Dynamic Environments
Computational Experiments
Experimental Design
Results and Discussion
Conclusion and Future Work
References
Memory-Based Immigrants for Ant Colony Optimization in Changing Environments
Introduction
DTSP with Cyclic Traffic Jams
ACO for the DTSP
Standard ACO
Population-Based ACO
Memory-Based Immigrants ACO for the DTSP
Simulation Experiments
Experimental Setup
Experimental Results and Analysis
Conclusions
References
Solving Dynamic Constrained Optimization Problems with Asynchronous Change Pattern
Introduction
Dynamic Fitness Landscapes and Dynamic Constraints
NumericalExperiments
Consider Schemes
The Benchmark and Its Properties
Recording Performance
Concluding Remarks
References
Flexible Variable Neighborhood Search in Dynamic Vehicle Routing
Introduction
Background
Problem Description
Static VRP
VRP with Dynamic Requests
FlexibleVNSforVRPDR
Algorithm Description
Flexible Management of Constraints
Experimental Study
Numerical Results
Alpha Parameter
Flexibility vs. Multi-populations
Conclusion
References
CHC-Based Algorithms for the Dynamic Traveling Salesman Problem
Introduction
The Dynamic Traveling Salesman Problem
The CHC Algorithms
Experimental Design
Dynamic TSP
Parameters Setting
Results
Conclusions and Future Work
References
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken