Lifelong Machine Learning

For more info about our services contact : help@bestpfe.com

Table of contents

1 Introduction 
1.1 Context
1.2 Autonomous Robots, Agents, and Swarms
1.3 Outline of this Thesis
Part I Methodological Background 
2 Machine Learning and Adaptation 
2.1 Machine Learning for Robot Control
2.2 Offline Learning vs. Online Adaptation
2.3 Adaptation to Several Tasks
2.4 Conclusion
3 Evolutionary Robotics 
3.1 Evolutionary Algorithms for Robot Control
3.2 Evolutionary Swarm Robotics
3.3 Embodied Evolutionary Robotics
3.4 Conclusion
Part II Distributed Adaptation of Swarm Robot Behavior
4 Approach for Swarm Robot Adaptation 
4.1 Baseline Algorithm: vanilla EE
4.2 Adaptation to Changing Conditions
4.3 Research Questions and Contributions
5 Influence of Selection Pressure in Distributed EER 
5.1 Exploration vs. Exploitation Dilemma
5.2 EER Algorithm and Selection Operators
5.3 Experiments
5.4 Conclusion
6 Collaborative Item Collection using Embodied Evolution 
6.1 Evolving Collaborative Behavior
6.2 Collaborative Item Collection
6.3 Experiments
6.4 Results and Analysis
6.5 Conclusion
7 Augmenting Neural Controllers in Embodied Evolution 
7.1 Motivation
7.2 Evolution of Neural Topologies
7.3 odNEAT with Logical Markers
7.4 Experiments
7.5 Conclusion
8 Evolutionary Adaptation to Changing Conditions 
8.1 Adaptation to Changing Conditions
8.2 Forgetting-Avoidance Evolutionary Algorithm
8.3 Preliminary Experiments
8.4 Toward Forgetting Avoidance in Distributed EER
8.5 Conclusion
9 Conclusion and Perspectives 
9.1 Summary
9.2 Contributions
9.3 Perspectives
List of Publications
Bibliography