From black and white to full color: extending redescription mining outside the Boolean world

Abstract

Redescription mining is a powerful data analysis tool that is used to find multiple descriptions of the same entities. Consider geographical regions as an example. They can be characterized by the fauna that inhabits them on one hand and by their meteorological conditions on the other hand. Finding such redescriptors, a task known as niche‐finding, is of much importance in biology. Current redescription mining methods cannot handle other than Boolean data. This restricts the range of possible applications or makes discretization a pre‐requisite, entailing a possibly harmful loss of information. In niche‐finding, while the fauna can be naturally represented using a Boolean presence/absence data, the weather cannot. In this paper, we extend redescription mining to categorical and real‐valued data with possibly missing values using a surprisingly simple and efficient approach. We provide extensive experimental evaluation to study the behavior of the proposed algorithm. Furthermore, we show the statistical significance of our results using recent innovations on randomization methods. © 2012 Wiley Periodicals, Inc. Statistical Analysis and Data Mining, 2012

Local Subgroup Discovery for Eliciting and Understanding New Structure-Odor Relationships

“…Indeed, we recall after the two most general approaches that can deal with our problem although only partially, namely subgroup discovery [14] and redescription mining [9]. We demonstrate this fact and highlight their weaknesses through an application example.…”

“…We need thus to consider not only all the possible subgroups, but all label subsets for each subgroup. In other settings, this search space is actually considered by a method called Redescription Mining [9].…”

“…Redescription mining (RM) [9] aims at finding two ways of describing a same set of objects. For that, two datasets are given with different attributes but the same set of object ID.…”

“…-For (i), we consider the enumeration of pairs consisting of a description and a subset of labels (a variant of redescription mining [9]). -For (ii), we propose a new measure derived from the F-score but less skewed by imbalance distribution of labels and that can dynamically consider the label distributions.…”

Computational biomes: The ecometrics of large mammal teeth

“…Others learn predictive models, namely classification trees, from which queries are then extracted (Ramakrishnan et al, 2004;Zinchenko et al, 2015). Yet others rely on empirically engineered searches to build the queries step by step (Gallo et al, 2008;Galbrun and Miettinen, 2012). In this study, we used the REREMI algorithm (Galbrun and Miettinen, 2012) for obtaining redescriptions.…”