Integrating memetic search into the BioHEL evolutionary learning system for large-scale datasets

Abstract

Local search methods are widely used to improve the performance of evolutionary computation algorithms in all kinds of domains. Employing advanced and efficient exploration mechanisms becomes crucial in complex and very large (in terms of search space) problems, such as when employing evolutionary algorithms to large-scale data mining tasks. Recently, the GAssist Pittsburgh evolutionary learning system was extended with memetic operators for discrete representations that use information from the supervised learning process to heuristically edit classification rules and rule sets. In this paper we first adapt some of these operators to BioHEL, a different evolutionary learning system applying the iterative learning approach, and afterwards propose versions of these operators designed for continuous attributes and for dealing with noise. The performance of all these operators and their combination is extensively evaluated on a broad range of synthetic large-scale datasets to identify the settings that present the best balance between efficiency and accuracy. Finally, the identified best configurations are compared with other classes of machine learning methods on both synthetic and real-world large-scale datasets and show very competent performance.

Mechanisms to Alleviate Over-Generalization in XCS for Continuous-Valued Input Spaces

“…XCSR only used explore steps in combination with an -decay action selection regime, parameterized as follows: = 1.0, fin = 0.02 and a decay fraction of 10% of the episodes, i.e., in the reported experiments the was decayed from 1.0 to 0.02 over the first 1000 episodes. GridWorld(0.07) and GridWorld(0.05) shared the same parameterization 8 , based on the settings for Puddles (0.1) in [19]. The conditions of the classifiers were encoded by unordered bound hyperrectangular 8 Analogous to 6-RMP, except: N = 10,000, = 0.95, mna = 4, GA = 50, del = 50, sub = 50, 0 = 0.005, m 0 = 0.25, r 0 = 0.5.…”

“…GridWorld(0.07) and GridWorld(0.05) shared the same parameterization 8 , based on the settings for Puddles (0.1) in [19]. The conditions of the classifiers were encoded by unordered bound hyperrectangular 8 Analogous to 6-RMP, except: N = 10,000, = 0.95, mna = 4, GA = 50, del = 50, sub = 50, 0 = 0.005, m 0 = 0.25, r 0 = 0.5. 7 Analogous to 6-RMP, except: N = 6400, GA = 48, del = 50, sub = 50, 0 = 1.0, red = 1.0, m 0 = 0.5, r 0 = 1.0.…”

“…In addition to the previously mentioned procedures developed especially for XCS with binary inputs, there are further specialization mechanisms in other LCSs. In the field of evolutionary learning systems, rule specialization methods have also been introduced, such as the so-called memetic operators in BioHEL [8]. The Mutespec operator [10], introduced for the ALECSYS classifier system, bases the identification of over-generals on the variance of the rewards received by a classifier.…”

A Tandem Evolutionary Algorithm for Identifying Causal Rules from Complex Data

“…Unfortunately, few available benchmark problems exist in the literature with tunable heterogeneity, epistasis, and overlap, making it challenging to test the sensitivity of the algorithm to each of these features. While others have used k-DNF functions as benchmarks that include heterogeneity, epistasis, and overlap Franco et al, 2012;Calian and Bacardit, 2013), the random way in which these problems are generated does not allow systematic control on the degree of overlap or class imbalance. After spending significant time trying to generate custom benchmark problems (e.g., Hanley et al, 2016), we appreciate the difficulty in designing appropriate tunable benchmarks.…”

“…The multiplexer problem, designed to predict the output of an electronic multiplexer circuit, is another scalable Boolean benchmark problem. The multiplexer problem was first introduced to the machine learning community by Barto (1985), and has been a standard benchmark problem for testing LCS approaches for decades (Wilson, 1987a,b;Booker, 1989;Goldberg, 1989;De Jong and Spears, 1991;Butz et al, 2003Butz et al, , 2004Butz et al, , 2005Bacardit and Krasnogor, 2006;Llorà et al, 2008;Franco et al, 2011;Ioannides et al, 2011;Calian and Bacardit, 2013;Iqbal et al, 2012Iqbal et al, , 2013aIqbal et al, ,b,c, 2014Iqbal et al, , 2015Urbanowicz and Moore, 2015).…”