Software Engineering · Sources
Technical Reading List
Search, MOEAs, and GALE. A separate shelf from the Philosopher's Pot reading list — journals, the dissertation, and the classic baselines behind Search Engineering.
Faster Evolutionary Multi-Objective Optimization via GALE, the Geometric Active LearnerJoseph Krall, 2014
PhD dissertation, West Virginia University
The long form of GALE: FastMap splits, continuous domination, mutate toward heaven, and why thrift beats shotgun sampling when each evaluation hurts.
In Search Engineering · Source
GALE: Geometric Active Learning for Search-Based Software EngineeringJoseph Krall, Tim Menzies, Misty Davies, 2015
IEEE Transactions on Software Engineering, 41(10)
The archival GALE paper. Comparable fronts to NSGA-II and SPEA2 in dozens of evaluations, not thousands.
In Search Engineering · DOI
Learning Mitigations for Pilot Issues When Landing Aircraft (via Multiobjective Optimization and Multiagent Simulations)Joseph Krall, Tim Menzies, Misty Davies, 2016
IEEE Transactions on Human-Machine Systems, 46(2)
GALE on the NASA Ames / Georgia Tech pilot-task model — expensive runs, mitigations found in minutes instead of hours.
In Search Engineering · DOI
Learning the Task Management Space of an Aircraft Approach ModelJoseph Krall, Tim Menzies, Misty Davies, 2014
AAAI Spring Symposium; NASA NTRS
Earlier cut of the approach-model study: GALE exploring cognitive structures in the WMC simulation.
In Search Engineering · NTRS
A Fast and Elitist Multiobjective Genetic Algorithm: NSGA-IIKalyanmoy Deb, Amrit Pratap, Sameer Agarwal, T. Meyarivan, 2002
IEEE Transactions on Evolutionary Computation, 6(2)
Non-dominated sorting plus crowding. Still the default "genetic algorithm" people mean for two-objective problems.
In Search Engineering · DOI
SPEA2: Improving the Strength Pareto Evolutionary AlgorithmEckart Zitzler, Marco Laumanns, Lothar Thiele, 2001
TIK-Report 103, ETH Zurich
Strength Pareto with an archive and density estimation — the other classic baseline beside NSGA-II.
In Search Engineering · DOI
MOEA/D: A Multiobjective Evolutionary Algorithm Based on DecompositionQingfu Zhang, Hui Li, 2007
IEEE Transactions on Evolutionary Computation, 11(6)
Decompose the front into neighboring scalar subproblems. The decomposition lineage's workhorse.
In Search Engineering · DOI
An Evolutionary Many-Objective Optimization Algorithm Using Reference-Point-Based Nondominated Sorting Approach, Part IKalyanmoy Deb, Himanshu Jain, 2014
IEEE Transactions on Evolutionary Computation, 18(4)
NSGA-III — reference points so many-objective search still spreads across the front.
In Search Engineering · DOI
FastMap: A Fast Algorithm for Indexing, Data-Mining and Visualization of Traditional and Multimedia DatasetsChristos Faloutsos, King-Ip Lin, 1995
ACM SIGMOD
The near-linear projection trick under GALE's WHERE splits — poles, a line, order the cloud.
In Search Engineering · DOI
Genetic Algorithms for Multiobjective Optimization: Formulation, Discussion and GeneralizationCarlos M. Fonseca, Peter J. Fleming, 1993
Proceedings of the Fifth International Conference on Genetic Algorithms
Early MOGA ranking and niching — and the Fonseca-Fleming test function that traps directed geometric search.
In Search Engineering · PDF
Comparison of Multiobjective Evolutionary Algorithms: Empirical ResultsEckart Zitzler, Kalyanmoy Deb, Lothar Thiele, 2000
Evolutionary Computation, 8(2)
The ZDT suite (ZDT1, ZDT3, and friends) that still anchors fair MOEA races — including the demo above.
In Search Engineering · DOI
Particle Swarm OptimizationJames Kennedy, Russell Eberhart, 1995
IEEE International Conference on Neural Networks
Bird-flock search as an algorithm — the wildlife rhyme in the opening of Search Engineering.
In Search Engineering · DOI
Ant System: Optimization by a Colony of Cooperating AgentsMarco Dorigo, Vittorio Maniezzo, Alberto Colorni, 1996
IEEE Transactions on Systems, Man, and Cybernetics, Part B, 26(1)
Ant trails as search — the other wildlife rhyme beside particle swarms.
In Search Engineering · DOI