We present an efficient feature selection method that can find all multiplicative combinations of continuous features that are statistically significantly associated with the class variable, while rigorously correcting for multiple testing. The key to overcome the combinatorial explosion in the number of candidates is to derive a lower bound on the $p$-value for each feature combination, which enables us to massively prune combinations that can never be significant and gain more statistical power. While this problem has been addressed for binary features in the past, we here present the first solution for continuous features. In our experiments, our novel approach detects true feature combinations with higher precision and recall than competing methods that require a prior binarization of the data.
Significant pattern mining, the problem of finding itemsets that are significantly enriched in one class of objects, is statistically challenging, as the large space of candidate patterns leads to an enormous multiple testing problem. Recently, the concept of testability was proposed as one approach to correct for multiple testing in pattern mining while retaining statistical power. Still, these strategies based on testability do not allow one to condition the test of significance on the observed covariates, which severely limits its utility in biomedical applications. Here we propose a strategy and an efficient algorithm to perform significant pattern mining in the presence of categorical covariates with K states.
Finding statistically significant interactions between binary variables is computationally and statistically challenging in high-dimensional settings, due to the combinatorial explosion in the number of hypotheses. Terada et al. recently showed how to elegantly address this multiple testing problem by excluding non-testable hypotheses. Still, it remains unclear how their approach scales to large datasets. We here proposed strategies to speed up the approach by Terada et al. and evaluate them thoroughly in 11 real-world benchmark datasets. We observe that one approach, incremental search with early stopping, is orders of magnitude faster than the current state-of-the-art approach.
The problem of finding itemsets that are statistically significantly enriched in a class of transactions is complicated by the need to correct for multiple hypothesis testing. Pruning untestable hypotheses was recently proposed as a strategy for this task of significant itemset mining. It was shown to lead to greater statistical power, the discovery of more truly significant itemsets, than the standard Bonferroni correction on real-world datasets. An open question, however, is whether this strategy of excluding untestable hypotheses also leads to greater statistical power in subgraph mining, in which the number of hypotheses is much larger than in itemset mining. Here we answer this question by an empirical investigation on eight popular graph benchmark datasets. We propose a new efficient search strategy, which always returns the same solution as the state-of-the-art approach and is approximately two orders of magnitude faster. Moreover, we exploit the dependence between subgraphs by considering the effective number of tests and thereby further increase the statistical power.
Apr 01 2013 cs.CV
Methodological contributions: This paper introduces a family of kernels for analyzing (anatomical) trees endowed with vector valued measurements made along the tree. While state-of-the-art graph and tree kernels use combinatorial tree/graph structure with discrete node and edge labels, the kernels presented in this paper can include geometric information such as branch shape, branch radius or other vector valued properties. In addition to being flexible in their ability to model different types of attributes, the presented kernels are computationally efficient and some of them can easily be computed for large datasets (N of the order 10.000) of trees with 30-600 branches. Combining the kernels with standard machine learning tools enables us to analyze the relation between disease and anatomical tree structure and geometry. Experimental results: The kernels are used to compare airway trees segmented from low-dose CT, endowed with branch shape descriptors and airway wall area percentage measurements made along the tree. Using kernelized hypothesis testing we show that the geometric airway trees are significantly differently distributed in patients with Chronic Obstructive Pulmonary Disease (COPD) than in healthy individuals. The geometric tree kernels also give a significant increase in the classification accuracy of COPD from geometric tree structure endowed with airway wall thickness measurements in comparison with state-of-the-art methods, giving further insight into the relationship between airway wall thickness and COPD. Software: Software for computing kernels and statistical tests is available at http://image.diku.dk/aasa/software.php.
Motivation: The rapid growth in genome-wide association studies (GWAS) in plants and animals has brought about the need for a central resource that facilitates i) performing GWAS, ii) accessing data and results of other GWAS, and iii) enabling all users regardless of their background to exploit the latest statistical techniques without having to manage complex software and computing resources. Results: We present easyGWAS, a web platform that provides methods, tools and dynamic visualizations to perform and analyze GWAS. In addition, easyGWAS makes it simple to reproduce results of others, validate findings, and access larger sample sizes through merging of public datasets. Availability: Detailed method and data descriptions as well as tutorials are available in the supplementary materials. easyGWAS is available at http://easygwas.tuebingen.mpg.de/. Contact: firstname.lastname@example.org
Jun 23 2009 cs.LG
In this paper, we present two classes of Bayesian approaches to the two-sample problem. Our first class of methods extends the Bayesian t-test to include all parametric models in the exponential family and their conjugate priors. Our second class of methods uses Dirichlet process mixtures (DPM) of such conjugate-exponential distributions as flexible nonparametric priors over the unknown distributions.
Jul 02 2008 cs.LG
We present a unified framework to study graph kernels, special cases of which include the random walk graph kernel \citepGaeFlaWro03,BorOngSchVisetal05, marginalized graph kernel \citepKasTsuIno03,KasTsuIno04,MahUedAkuPeretal04, and geometric kernel on graphs \citepGaertner02. Through extensions of linear algebra to Reproducing Kernel Hilbert Spaces (RKHS) and reduction to a Sylvester equation, we construct an algorithm that improves the time complexity of kernel computation from $O(n^6)$ to $O(n^3)$. When the graphs are sparse, conjugate gradient solvers or fixed-point iterations bring our algorithm into the sub-cubic domain. Experiments on graphs from bioinformatics and other application domains show that it is often more than a thousand times faster than previous approaches. We then explore connections between diffusion kernels \citepKonLaf02, regularization on graphs \citepSmoKon03, and graph kernels, and use these connections to propose new graph kernels. Finally, we show that rational kernels \citepCorHafMoh02,CorHafMoh03,CorHafMoh04 when specialized to graphs reduce to the random walk graph kernel.
We propose a framework for analyzing and comparing distributions, allowing us to design statistical tests to determine if two samples are drawn from different distributions. Our test statistic is the largest difference in expectations over functions in the unit ball of a reproducing kernel Hilbert space (RKHS). We present two tests based on large deviation bounds for the test statistic, while a third is based on the asymptotic distribution of this statistic. The test statistic can be computed in quadratic time, although efficient linear time approximations are available. Several classical metrics on distributions are recovered when the function space used to compute the difference in expectations is allowed to be more general (eg. a Banach space). We apply our two-sample tests to a variety of problems, including attribute matching for databases using the Hungarian marriage method, where they perform strongly. Excellent performance is also obtained when comparing distributions over graphs, for which these are the first such tests.
Apr 23 2007 cs.LG
We introduce a framework for filtering features that employs the Hilbert-Schmidt Independence Criterion (HSIC) as a measure of dependence between the features and the labels. The key idea is that good features should maximise such dependence. Feature selection for various supervised learning problems (including classification and regression) is unified under this framework, and the solutions can be approximated using a backward-elimination algorithm. We demonstrate the usefulness of our method on both artificial and real world datasets.