(Downloads - 0)
For more info about our services contact : help@bestpfe.com
Table of contents
I Introduction
1 Introduction
1.1 Motivation
1.2 Automated Algorithm Selection
1.3 Our Key Findings
1.4 Thesis Outline
II Landscape-Aware Selection of Metaheuristics
2 Continuous Black-Box Optimization Algorithms
2.1 Black-Box Optimization
2.2 Evolutionary Computation
2.3 The Covariance Matrix Adaptation Evolution Strategy (CMA-ES)
2.4 Dierential Evolution (DE)
2.5 Particle Swarm Optimization (PSO)
2.6 Direct Search Methods
2.7 L-BFGS-B Algorithm
2.8 Automated Algorithm Conguration
3 Landscape-Aware Algorithm Selection
3.1 Motivation
3.2 Landscape-Aware Algorithm Selection
3.3 Landscape-Aware Algorithm Selection Pipeline
4 Characterizing Problem Instances via Landscape Features
4.1 Fitness Landscape Analysis
4.2 Exploratory Landscape Analysis
III Analysis of Landscape Features
5 Exploratory Landscape Features Properties
5.1 Design of Experiments
5.2 Stability
5.3 In uence of Sampling Strategy
5.4 Expressiveness
5.5 Robustness
5.6 Invariance to Transformations
5.7 Sensitivity to Noise
5.8 Discussion
IV Optimization of Radar Networks
6 Background on Radar Operation
6.1 Introduction
6.2 History of Radar Development
6.3 Basic Principle
6.4 Radar Equation
6.5 Radar Cross Section
6.6 Swerling Models
6.7 Probability of Detection
7 Radar Network Modeling
7.1 gis: Radar Network Modeling Framework
7.2 Target Characteristics
7.3 Radar Models and Parameters
7.4 Radar Network Use-Cases
7.5 Thesis Use-Cases
7.6 Geographical Data
8 Solving the Radar Network Conguration Problem
8.1 Problem Instances
8.2 Algorithm Portfolio
8.3 Experimental Setup
8.4 Results for the Unconstrained Use-Case
8.5 Results for the Constrained Use-Case
8.6 Comparison with Manual Optimization
9 Landscape-Aware Algorithm Selection on the Unconstrained Use-Case
9.1 Design of the Selector
9.2 Problem Characteristics
9.3 Denition of the SBS
9.4 Selector Performances
9.5 Discussion
V Conclusions
Conclusions
10.1 Summary of Contributions
10.2 Perspectives
A Summary of Papers and Industrial Achievements
A.1 Academic Papers and Presentations
A.2 Industrial Achievements
B Best Performing Algorithm for each Instance
B.1 Best Performing Algorithm in Median
B.2 Best Performing Algorithm for the 2% Quantile
C Radar Network Conguration Contest
C.1 Radar Network Conguration Contest Results
C.2 Radar Network Conguration Contest Poster
