An analysis pipeline with statistical and visualization-guided knowledge discovery for Michigan-style learning classifier systems

Abstract

Michigan-style learning classifier systems (M-LCSs) represent an adaptive and powerful class of evolutionary algorithms which distribute the learned solution over a sizable population of rules. However their application to complex real world data mining problems, such as genetic association studies, has been limited. Traditional knowledge discovery strategies for M-LCS rule populations involve sorting and manual rule inspection. While this approach may be sufficient for simpler problems, the confounding influence of noise and the need to discriminate between predictive and non-predictive attributes calls for additional strategies. Additionally, tests of significance must be adapted to M-LCS analyses in order to make them a viable option within fields that require such analyses to assess confidence. In this work we introduce an M-LCS analysis pipeline that combines uniquely applied visualizations with objective statistical evaluation for the identification of predictive attributes, and reliable rule generalizations in noisy single-step data mining problems. This work considers an alternative paradigm for knowledge discovery in M-LCSs, shifting the focus from individual rules to a global, population-wide perspective. We demonstrate the efficacy of this pipeline applied to the identification of epistasis (i.e., attribute interaction) and heterogeneity in noisy simulated genetic association data.

An Extended Michigan-Style Learning Classifier System for Flexible Supervised Learning, Classification, and Data Mining

“…Upon request, ExSTraCS will yield up to four distinct output files after the final iteration, or any iteration at which a full evaluation is requested. These include (1) the population of classifiers collectively constituting the prediction 'model' (Note: ExSTraCS is uniquely set up to load a given population file and continue learning from where it left off), (2) population statistics, summarizing major performance statistics including global training and testing accuracy of the classifier population [12], (3) co-occurrence scores for the top specified pairs of attributes in the dataset [12], and (4) attribute tracking scores for each instance in the dataset [7]. These outputs may be evaluated and visualized to facilitate knowledge discovery as described in [12].…”

“…Specifically, classifier conditions can only specify one state for discrete attributes, while GABIL allows for some subset of attribute states to be simultaneously specified. While this may be advantageous for evolving a maximally compact rule-set, this approach is not in-line with the global approach to knowledge discovery proposed in [12] which relies on important attributes being specified more often across rules in the greater population, a valuable component to addressing significant noise in LCS data mining. Additionally, this stricter representation yields individual rules that are arguably easier to interpret (less ambiguity within the IF/THEN statement) and that are likely be more accurate individual predictors (since they independently capture a more specific set of attribute states).…”

“…Rule compaction utilizes the whole training dataset to consolidate the classifier population with the goal of improving interpretation and knowledge discovery. Comparisons in [10] suggested that simple Quick Rule Filtering (QRF) was both the fastest, and particularly was well suited to the theme of global knowledge discovery [12] where it is more important to preserve or improve performance than to minimize rule population size(useful for knowledge discovery by manual rule inspection) [10]. This study applies QRF.…”

Hard Data Analytics Problems Make for Better Data Analysis Algorithms: Bioinformatics as an Example

“…Other researchers have applied similar techniques (in terms of both rule-based machine learning and knowledge extraction) to synthetic and real human SNP data. [29][30][31] Toward Knowledge Intensive Biodata Mining Through these four case studies, we have illustrated the central role that data mining has in the analysis of biological data and the variety of challenging mining tasks that such analysis requires. We also demonstrated the use of white-box data mining techniques such as rulebased machine learning or genetic programming, coupled with sophisticated information visualization techniques, to discover new knowledge from the data.…”

Efficient training set use for blood pressure prediction in a large scale learning classifier system

“…Some, like EC-Star, distribute the algorithm. For example, Urbanowicz et al [23] use a cluster and Franco et al [6] use a GPU. EC-Star's use of volunteer compute nodes appears to be unique and allows many more resources (in the hundreds or thousands) to be enlisted.…”