Contrast Set Mining Research Articles

A contrast set mining based approach for cancer subtype analysis

The task of detecting common and unique characteristics among different cancer subtypes is an important focus of research that aims to improve personalized therapies. Unlike current approaches mainly based on predictive techniques, our study aims to improve the knowledge about the molecular mechanisms that descriptively led to cancer, thus not requiring previous knowledge to be validated. Here, we propose an approach based on contrast set mining to capture high-order relationships in cancer transcriptomic data. In this way, we were able to extract valuable insights from several cancer subtypes in the form of highly specific genetic relationships related to functional pathways affected by the disease. To this end, we have divided several cancer gene expression databases by the subtype associated with each sample to detect which gene groups are related to each cancer subtype. To demonstrate the potential and usefulness of the proposed approach we have extensively analysed RNA-Seq gene expression data from breast, kidney, and colon cancer subtypes. The possible role of the obtained genetic relationships was further evaluated through extensive literature research, while its prognosis was assessed via survival analysis, finding gene expression patterns related to survival in various cancer subtypes. Some gene associations were described in the literature as potential cancer biomarkers while other results have been not described yet and could be a starting point for future research.

Succinct contrast sets via false positive controlling with an application in clinical process redesign

Many applications of intelligent systems involve understanding a group of contrastively different outcome (e.g., all survivors of a deadly cancer, a top performing team in a large corporation). The intelligent system needs to identify attributes (features) which best describe or explain the group versus its alternatives. In data mining, this problem is studied under the framework of contrast set mining (CSM). Although CSM is not new, the era of big data has produced new computational and statistical challenges. In particular, existing algorithms fail (1) to perform efficiently in terms of runtime on large-scale datasets and (2) to accommodate simultaneous inference on an overwhelming array of features which are often repetitive and collinear. In this paper, we develop a CSM algorithm which addresses both challenges. The computational challenge is addressed with a tree structure and two theorems while the statistical challenge is addressed with the application of false discovery rate for multiple testing. The computational and statistical advantages of the proposed algorithm over three state-of-the-art algorithms are demonstrated with comprehensive experiments. In addition, we also show the effectiveness of our proposed method in an intelligence-system application involving hospital process redesign. The proposed method not only improves the performance of machine learning systems, but also generates succinct and insightful patterns directly relevant to clinical decision-making.

Supervised descriptive pattern discovery in Native American music

The discovery of recurrent patterns in groups of songs is an important first step in computational corpus analysis. In this paper, computational techniques of supervised descriptive pattern discovery are applied to model and extend ethnomusicological analyses of Native American music. Using a corpus of over 2000 songs collected and transcribed by anthropologist Frances Densmore and building on Densmore’s own music content features, the analysis identifies musical differences between indigenous groups and between musical style areas of the North American continent. Contrast set mining is adapted to discover global-feature patterns which are distinctive for a group, statistically significant and maximally general. The work extends previous descriptive studies in computational folk music analysis by considering feature-set patterns of variable size. Discovered patterns confirm, differentiate and complement ethnomusicological observations on Native American music.

Time series joins, motifs, discords and shapelets: a unifying view that exploits the matrix profile

The last decade has seen a flurry of research on all-pairs-similarity-search (or similarity joins) for text, DNA and a handful of other datatypes, and these systems have been applied to many diverse data mining problems. However, there has been surprisingly little progress made on similarity joins for time series subsequences. The lack of progress probably stems from the daunting nature of the problem. For even modest sized datasets the obvious nested-loop algorithm can take months, and the typical speed-up techniques in this domain (i.e., indexing, lower-bounding, triangular-inequality pruning and early abandoning) at best produce only one or two orders of magnitude speedup. In this work we introduce a novel scalable algorithm for time series subsequence all-pairs-similarity-search. For exceptionally large datasets, the algorithm can be trivially cast as an anytime algorithm and produce high-quality approximate solutions in reasonable time and/or be accelerated by a trivial porting to a GPU framework. The exact similarity join algorithm computes the answer to the time series motif and time series discord problem as a side-effect, and our algorithm incidentally provides the fastest known algorithm for both these extensively-studied problems. We demonstrate the utility of our ideas for many time series data mining problems, including motif discovery, novelty discovery, shapelet discovery, semantic segmentation, density estimation, and contrast set mining. Moreover, we demonstrate the utility of our ideas on domains as diverse as seismology, music processing, bioinformatics, human activity monitoring, electrical power-demand monitoring and medicine.

Handling Sparse Data Sets by Applying Contrast Set Mining in Feature Selection

A data set is sparse if the number of samples in a data set is not sufficient to model the data accurately. Recent research emphasized interest in applying data mining and feature selection techniques to real world problems, many of which are characterized as sparse data sets. The purpose of this research is to define new techniques for feature selection in order to improve classification accuracy and reduce the time required for feature selection on sparse data sets. The extensive comparison with benchmarking feature selection techniques on 64 sparse data sets was conducted. Results have shown superiority of contrast set mining techniques in more than 80% of the analysis on sparse data sets. This paper provides a study on the new methodologies and detected superiority in handling data sparsity.

DisCoSet: Discovery of Contrast Sets to Reduce Dimensionality and Improve Classification

AbstractTraditionally, contrast set mining aims at finding a set of rules that best distinguish the instances of different user-defined groups. Contrast sets are conjunctions of attribute-value pairs that are significantly more frequent in one group than in other groups. Typically, these contrast sets are extracted from categorical data or discretized numerical data. Existing methods of rule-based contrast sets require some user-defined thresholds to select the contrast sets. In this paper, we propose a greedy algorithm, called DisCoSet, to find incrementally a minimum set of local features that best distinguishes a class from other classes without resorting to discretization. The discovered contrast sets reduce the dimensionality of the feature vectors considerably and improve the classification accuracy significantly. We show that the proposed algorithm reduces the dimensionality of class instances by 40%-97% of the original length and yet improves classification accuracy by 10%-24% using different type...

Contrast set mining in temporal databases

AbstractUnderstanding the underlying differences between groups or classes in certain contexts can be of the utmost importance. Contrast set mining relies on discovering significant patterns by contrasting two or more groups. A contrast set is a conjunction of attribute–value pairs that differ meaningfully in its distribution across groups. A previously proposed technique is rules for contrast sets, which seeks to express each contrast set found in terms of rules. This work extends rules for contrast sets to a temporal data mining task. We define a set of temporal patterns in order to capture the significant changes in the contrasts discovered along the considered time line. To evaluate the proposal accuracy and ability to discover relevant information, two different real‐life data sets were studied using this approach.

A Tree-Based Contrast Set-Mining Approach to Detecting Group Differences

Understanding differences between groups in a data set is one of the fundamental tasks in data analysis. As relevant applications accumulate, data-mining methods have been developed to specifically address the problem of group difference detection. Contrast set mining discovers group differences in the form of conjunction of feature-value pairs or items. In this paper, we incorporate absolute difference, relative difference, and statistical significance in our definition of a group difference, and develop a novel method named DIFF that uses the prefix-tree structure to compress the search space, follows a tree traversal procedure to discover the complete set of significant group differences, and employs efficient pruning strategies to expedite the search process. We conducted comprehensive experiments to compare our method with existing methods on completeness of results, pruning efficiency, and computational efficiency. The experiments demonstrate that our method guarantees completeness of results and achieves higher pruning efficiency and computational efficiency compared to STUCCO. In addition, our definition of group difference is more general than STUCCO. Our method is more effective than traditional approaches, such as classification trees, in discovering the complete set of significant group differences.

Supervised Descriptive Rule Discovery: A Unifying Survey of Contrast Set, Emerging Pattern and Subgroup Mining

This paper gives a survey of contrast set mining (CSM), emerging pattern mining (EPM), and subgroup discovery (SD) in a unifying framework named supervised descriptive rule discovery. While all these research areas aim at discovering patterns in the form of rules induced from labeled data, they use different terminology and task definitions, claim to have different goals, claim to use different rule learning heuristics, and use different means for selecting subsets of induced patterns. This paper contributes a novel understanding of these subareas of data mining by presenting a unified terminology, by explaining the apparent differences between the learning tasks as variants of a unique supervised descriptive rule discovery task and by exploring the apparent differences between the approaches. It also shows that various rule learning heuristics used in CSM, EPM and SD algorithms all aim at optimizing a trade off between rule coverage and precision. The commonalities (and differences) between the approaches are showcased on a selection of best known variants of CSM, EPM and SD algorithms. The paper also provides a critical survey of existing supervised descriptive rule discovery visualization methods.

CSM-SD: Methodology for contrast set mining through subgroup discovery

This paper addresses a data analysis task, known as contrast set mining, whose goal is to find differences between contrasting groups. As a methodological novelty, it is shown that this task can be effectively solved by transforming it to a more common and well-understood subgroup discovery task. The transformation is studied in two learning settings, a one-versus-all and a pairwise contrast set mining setting, uncovering the conditions for each of the two choices. Moreover, the paper shows that the explanatory potential of discovered contrast sets can be improved by offering additional contrast set descriptors, called the supporting factors. The proposed methodology has been applied to uncover distinguishing characteristics of two groups of brain stroke patients, both with rapidly developing loss of brain function due to ischemia:those with ischemia caused by thrombosis and by embolism, respectively.