For real: a thorough look at numeric attributes in subgroup discovery

Abstract

Subgroup discovery (SD) is an exploratory pattern mining paradigm that comes into its own when dealing with large real-world data, which typically involves many attributes, of a mixture of data types. Essential is the ability to deal with numeric attributes, whether they concern the target (a regression setting) or the description attributes (by which subgroups are identified). Various specific algorithms have been proposed in the literature for both cases, but a systematic review of the available options is missing. This paper presents a generic framework that can be instantiated in various ways in order to create different strategies for dealing with numeric data. The bulk of the work in this paper describes an experimental comparison of a considerable range of numeric strategies in SD, where these strategies are organised according to four central dimensions. These experiments are furthermore repeated for both the classification task (target is nominal) and regression task (target is numeric), and the strategies are compared based on the quality of the top subgroup, and the quality and redundancy of the top-k result set. Results of three search strategies are compared: traditional beam search, complete search, and a variant of diverse subgroup set discovery called cover-based subgroup selection. Although there are various subtleties in the outcome of the experiments, the following general conclusions can be drawn: it is often best to determine numeric thresholds dynamically (locally), in a fine-grained manner, with binary splits, while considering multiple candidate thresholds per attribute.

Data analytics accelerates the experimental discovery of new thermoelectric materials with extremely high figure of merit

“…Before discussing predictive analytic SISSO model for zT, we analyze qualitatively all measured properties contributing to thermoelectric performance using a data-mining approach subgroup discovery (SGD). [30][31][32][33] SGD nds statistically exceptional subgroups in a dataset described by statements (selectors) of the kind (feature 1 < ) AND (feature 2 > ) AND ... . The features include only temperature and experimentally known materials properties listed in Table 1 and Table S1.…”

Discovering outstanding subgroup lists for numeric targets using MDL

“…For the quality measure it uses the Weighted Kullback-Leibler without dispersion, i.e., W KL µ (s) = n s /σ d ( μd − μs ) 2 as described in Appendix 3, as the algorithm does not accept its dispersion-aware version used in Eq. (16).…”

“…The first reason for using greedy search to add one subgroup at the time, is its transparency, as it adds at each iteration the locally best subgroup found by the beam search. Beamsearch, on the other hand, was empirically shown, in the context of subgroup discovery for numeric targets, to be very competitive in terms of quality when compared to a complete search with an associated speedup improvement [16]. Also, its straightforward implementation allows to easily extend this framework to other types of targets, not just numeric.…”

Discovering Outstanding Subgroup Lists for Numeric Targets Using MDL

“…The first reason for using greedy search to add one subgroup at the time, is its transparency, as it adds at each iteration the locally best subgroup found by the beam search. Beam-search, on the other hand, was empirically shown, in the context of subgroup discovery for numeric targets, to be very competitive in terms of quality when compared to a complete search with an associated speedup improvement [14]. Also, its straightforward implementation allows to easily extend this framework to other types of targets, not just numeric.…”