Sporadic Model Building for Efficiency Enhancement of hBOA
Pelikan, M., Sastry, K., Goldberg, D. E. (2005)TR No.: 2005026 | Download PDF | Download PS
Abstract:
This paper describes and analyzes sporadic model building, which can be used to enhance the efficiency of the hierarchical Bayesian optimization algorithm (hBOA) and other estimation of distribution algorithms (EDAs). With sporadic model building, the structure of the probabilistic model is updated once every few iterations (generations), whereas in the remaining iterations only model parameters […]
Posted: May 12th, 2005 under Genetic algorithms. Comments: 1
Evaluation Consistency in iGAs: User Contradictions as Cycles in Partial-Ordering Graphs
Llorà, X., Alías, F. , Formiga, L., Sastry, K., Goldberg, D. E. (2005)TR No.: 2005022 | Download PDF | Download PS
Abstract:
Active interactive genetic algorithms (aiGAs) rely on actively optimizing synthetic fitness functions. In interactive genetic algorithms (iGAs) framework, user evaluations provide the necessary input for synthesizing a reasonably accurate surrogate fitness function that models user evaluations or, in other words, his/her decision preferences. User evaluations collected via tournament selection only provide partial-ordering relations between […]
Posted: May 8th, 2005 under Genetic algorithms. Comments: none
Fluctuating Crosstalk as a Source of Deterministic Noise and its Effects on GA Scalability
Sastry, K., Winward, P., Goldberg, D.E., Lima, C. (2005)TR No.: 2005025 | Download PDF | Download PS
Abstract:
This paper explores how fluctuating crosstalk in a
deterministic fitness function introduces noise into
genetic algorithms. We model fluctuating crosstalk
or nonlinear interactions among building blocks via
higher-order Walsh coefficients. The fluctuating
crosstalk behaves like exogenous noise and can be
handled by increasing the population size and run
duration. This behavior holds until the strength of
the crosstalk far exceeds […]
Posted: May 4th, 2005 under Genetic algorithms. Comments: none
Classifier Conditions based on Convex Hulls
Lanzi P.L., Wilson S.W. (2005)TR No.: 2005024 | Download PDF | Download PS
Abstract:
We introduce a novel representation of classifier conditions based on convex hulls. A classifier condition is represented by a sets of points in the problem space. These points identify a convex hull that delineates a convex region of the problem space. The condition matches all the problem instances inside such region. We apply XCSF with […]
Posted: April 24th, 2005 under Genetic algorithms. Comments: none
Generalization in XCSF for Real Inputs
Lanzi P.L., Loiacono D., Wilson S.W., Goldberg D.E. (2005)TR No.: 2005023 | Download PDF | Download PS
Abstract:
This report extends the analysis of generalization in XCSF with integer inputs (reported in the IlliGAL report 2005012) to the case of real inputs. The results we present here fully confirm the conclusions drawn from the previous analysis.
Related PostsGeneralization in the XCSF Classifier System: Analysis, Improvement, and ExtensionExtending XCSF Beyond Linear ApproximationXCS with Computed Prediction […]
Posted: April 20th, 2005 under Genetic algorithms. Comments: none
A Little Model of Optimal Group Size in Breakout Meetings
Goldberg, D.E. (2005)TR No.: 2005021 | Download PDF | Download PS
Abstract:
In 25 years of work in genetic algorithms (GAs), the
author has tried a variety of analytical
methods to better understand GAs and their
deriviatives. As outlined elsewhere (Goldberg,
2002),
that work initially focused on transform methods,
difference equations, and Markov chains, but
these tools were disappointing because they failed
to deliver results that were generally useful in
GA […]
Posted: April 16th, 2005 under Genetic algorithms. Comments: none
Multiobjective Learning Classifier Systems: An Overview
Bernadó-Mansilla, E., Llorà , X., Traus I. (2005)TR No.: 2005020 | Download PDF | Download PS
Abstract:
Learning concept descriptions from data is a complex multiobjective task.
The model induced by the learner should be accurate
so that it can represent precisely the data instances,
complete, which means it can be generalizable to new instances,
and minimum, or easily readable.
Learning Classifier Systems (LCSs) are a family or learners whose primary search
mechanism is a genetic algorithm.
Along the […]
Posted: April 12th, 2005 under Genetic algorithms. Comments: none
The Compact Classifier System: Scalability Analysis and First Results
Llorà, X., Sastry, K., Goldberg, D.E. (2005)TR No.: 2005019 | Download PDF | Download PS
Abstract:
This paper presents an analysis of how maximally general and
accurate rules can be evolved in a Pittsburgh-style classifier
system. In order to be able to perform such an analysis we introduce a
simple bare-bones Pittsburgh-style classifier systems—the compact classifier system (CCS)—based on estimation of distribution algorithms. Using a common rule encoding schemes of Pittsburgh-style classifier systems,
CCS mantains […]
Posted: April 8th, 2005 under Genetic algorithms. Comments: none
XCS with Computed Prediction in Continuous Multistep Environments
Lanzi P.L., Loiacono D., Wilson S.W., Goldberg D.E. (2005)TR No.: 2005018 | Download PDF | Download PS
Abstract:
We apply XCS with computed prediction (XCSF) to tackle multistep reinforcement learning problems involving continuous inputs. In essence we use XCSF as a method of generalized reinforcement learning. We show that in domains involving continuous inputs and delayed rewards XCSF can evolve compact populations of accurate maximally general classifiers which represent the optimal solution to […]
Posted: April 4th, 2005 under Genetic algorithms. Comments: none
Online Population Size Adjusting Using Noise and Substructural Measurements
Yu, T.-L., Sastry, K., Goldberg, D.E (2005)TR No.: 2005017 | Download PDF | Download PS
Abstract:
This paper proposes an online population size adjustment scheme for genetic algorithms. It utilizes linkage-model-building techniques to calculate the parameters used in facetwise population-sizing models. The methodology is demonstrated using
the dependency structure matrix genetic algorithm on a set of boundedly-difficult problems. Empirical results indicate that the
proposed method is both efficient and robust. If the […]
Posted: March 24th, 2005 under Genetic algorithms. Comments: none