Abstract

Abstract
Clustering is one of the most popular data mining techniques. In this article, we review the relevant methods and algorithms for designing cluster algorithms under the data streams computational model, and discuss research directions in tracking evolving clusters. © Springer Science+Business Media New York 2011, 2017

Abstract
Feature selection and feature ranking are two aspects of the same learning task. They are well studied in batch scenarios, but not in the streaming setting. This paper presents a study on feature ranking from data streams in online learning regression models. The main challenge here is the relevance of features might change over time: features relevant in the past might be irrelevant now and vice-versa. We propose three new online feature ranking algorithms designed for Hoeffding algorithms. We have implemented the three methods in AMRules, a streaming regression algorithm to learn model rules. We compare their behaviour experimentally and present the pros and cons of each method. Copyright 2017 ACM.

Abstract

Abstract

Abstract
The huge amount of genomic and transcriptomic data obtained to characterize human diversity can also be exploited to indirectly gather information on the human microbiome. Here we present the pipeline QmihR designed to identify and quantify the abundance of known microbiome communities and to search for new/rare pathogenic species in RNA-seq datasets. We applied QmihR to 36 RNA-seq tumor tissue samples from Ukrainian gastric carcinoma patients available in TCGA, in order to characterize their microbiome and check for efficiency of the pipeline. The microbes present in the samples were in accordance to published data in other European datasets, and the independent BLAST evaluation of microbiome-aligned reads confirmed that the assigned species presented the highest BLAST match-hits. QmihR is available at GitHub (https://github.com/ Pereira-lab/QmihR). © Springer International Publishing AG 2017.