Ensemble methods are a very diverse family of algorithms with a wide range of applications. One of the most commonly used is boosting, with the prominent Adaboost. Adaboost relies on greedily learning base classifiers that rectify the error from previous iterations. Then, it combines them through a weighted majority vote, based on their quality on the entire learning set. In this paper, we propose a supervised binary classification framework that propagates the local knowledge acquired during the boosting iterations to the prediction function. Based on this general framework, we introduce SamBA, an interpretable greedy ensemble method designed for fat datasets, with a large number of dimensions and a small number of samples. SamBA learns local classifiers and combines them, using a similarity function, to optimize its efficiency in data extraction. We provide a theoretical analysis of SamBA, yielding convergence and generalization guarantees. In addition, we highlight SamBA's empirical behavior in an extensive experimental analysis on both real biological and generated datasets, comparing it to state-of-the-art ensemble methods and similarity-based approaches.

We consider the kernel completion problem with the presence of multiple views in the data. In this context the data samples can be fully missing in some views, creating missing columns and rows to the kernel matrices that are calculated individually for each view. We propose to solve the problem of completing the kernel matrices with Cross-View Kernel Transfer (CVKT) procedure, in which the features of the other views are transformed to represent the view under consideration. The transformations are learned with kernel alignment to the known part of the kernel matrix, allowing for finding generalizable structures in the kernel matrix under completion. Its missing values can then be predicted with the data available in other views. We illustrate the benefits of our approach with simulated data, multivariate digits dataset and multi-view dataset on gesture classification, as well as with real biological datasets from studies of pattern formation in early Drosophila melanogaster embryogenesis.

SuMMIT ( Su pervised M ulti M odal I ntegration T ool) is a software offering many functionalities for running, tuning, and analyzing experiments of supervised classification tasks specifically designed for multi-view data sets. SuMMIT is part of a platform 1 that aggregates multiple tools to deal with multiview datasets such as scikit-multimodallearn (Benielli et al., 2021) or MAGE (Bauvin et al., 2021). This paper presents use cases of SuMMIT, including hyper-parameters optimization, demonstrating the usefulness of such a platform for dealing with the complexity of multi-view benchmarking on an imbalanced dataset. SuMMIT is powered by Python3 and based on scikit-learn , making it easy to use and extend by plugging one’s own specific algorithms, score functions or adding new features 2 . By using continuous integration, we encourage collaborative development.

scikit-multimodallearn is a Python library for multimodal supervised learning, licensed under Free BSD, and compatible with the well-known scikit-learn toolbox (Fabian Pedregosa, 2011). This paper details the content of the library, including a specific multimodal data formatting and classification and regression algorithms. Use cases and examples are also provided.

This paper explains our participation in task 1 of the CASE 2021 shared task. This task is about multilingual event extraction from news. We focused on sub-task 4, event information extraction. This sub-task has a small training dataset and we fine-tuned a multilingual BERT to solve this sub-task. We studied the instability problem on the dataset and tried to mitigate it.

Abstract Most epigenetic marks, such as Transcriptional Regulators or histone marks, are biological objects known to work together in n-wise complexes. A suitable way to infer such functional associations between them is to study the overlaps of the corresponding genomic regions. However, the problem of the statistical significance of n-wise overlaps of genomic features is seldom tackled, which prevent rigorous studies of n-wise interactions. We introduce OLOGRAM-MODL, which considers overlaps between n ≥ 2 sets of genomic regions, and computes their statistical mutual enrichment by Monte Carlo fitting of a Negative Binomial distribution, resulting in more resolutive P-values. An optional machine learning method is proposed to find complexes of interest, using a new itemset mining algorithm based on dictionary learning which is resistant to noise inherent to biological assays. The overall approach is implemented through an easy-to-use CLI interface for workflow integration, and a visual tree-based representation of the results suited for explicability. The viability of the method is experimentally studied using both artificial and biological data. This approach is accessible through the command line interface of the pygtftk toolkit, available on Bioconda and from https://github.com/dputhier/pygtftk

Background Accurate identification of Transcriptional Regulator binding locations is essential for analysis of genomic regions, including Cis Regulatory Elements. The customary NGS approaches, predominantly ChIP-Seq, can be obscured by data anomalies and biases which are difficult to detect without supervision. Results Here, we develop a method to leverage the usual combinations between many experimental series to mark such atypical peaks. We use deep learning to perform a lossy compression of the genomic regions’ representations with multiview convolutions. Using artificial data, we show that our method correctly identifies groups of correlating series and evaluates CRE according to group completeness. It is then applied to the ReMap database’s large volume of curated ChIP-seq data. We show that peaks lacking known biological correlators are singled out and less confirmed in real data. We propose normalization approaches useful in interpreting black-box models. Conclusion Our approach detects peaks that are less corroborated than average. It can be extended to other similar problems, and can be interpreted to identify correlation groups. It is implemented in an open-source tool called atyPeak.

Traditionally, kernel methods rely on the representer theorem which states that the solution to a learning problem is obtained as a linear combination of the data mapped into the reproducing kernel Hilbert space (RKHS). While elegant from theoretical point of view, the theorem is prohibitive for algorithms' scalability to large datasets, and the interpretability of the learned function. In this paper, instead of using the traditional representer theorem, we propose to search for a solution in RKHS that has a pre-image decomposition in the original data space, where the elements don't necessarily correspond to the elements in the training set. Our gradient-based optimisation method then hinges on optimising over possibly sparse elements in the input space, and enables us to obtain a kernel-based model with both primal and dual sparsity. We give theoretical justification on the proposed method's generalization ability via a Rademacher bound. Our experiments demonstrate a better scalability and interpretability with accuracy on par with the traditional kernel-based models.

Translating biomedical ontologies is an important challenge, but doing it manually requires much time and money. We study the possibility to use open-source knowledge bases to translate biomedical ontologies. We focus on two aspects: coverage and quality. We look at the coverage of two biomedical ontologies focusing on diseases with respect to Wikidata for 9 European languages (Czech, Dutch, English, French, German, Italian, Polish, Portuguese and Spanish) for both ontologies, plus Arabic, Chinese and Russian for the second one. We first use direct links between Wikidata and the studied ontologies and then use second-order links by going through other intermediate ontologies. We then compare the quality of the translations obtained thanks to Wikidata with a commercial machine translation tool, here Google Cloud Translation.

The $$\mathcal {C}$$ -bound is a tight bound on the true risk of a majority vote classifier that relies on the individual quality and pairwise disagreement of the voters and provides PAC-Bayesian generalization guarantees. Based on this bound, MinCq is a classification algorithm that returns a dense distribution on a finite set of voters by minimizing it. Introduced later and inspired by boosting, CqBoost uses a column generation approach to build a sparse $$\mathcal {C}$$ -bound optimal distribution on a possibly infinite set of voters. However, both approaches have a high computational learning time because they minimize the $$\mathcal {C}$$ -bound by solving a quadratic program. Yet, one advantage of CqBoost is its experimental ability to provide sparse solutions. In this work, we address the problem of accelerating the $$\mathcal {C}$$ -bound minimization process while keeping the sparsity of the solution and without losing accuracy. We present CB-Boost, a computationally efficient classification algorithm relying on a greedy–boosting-based– $$\mathcal {C}$$ -bound optimization. An in-depth analysis proves the optimality of the greedy minimization process and quantifies the decrease of the $$\mathcal {C}$$ -bound operated by the algorithm. Generalization guarantees are then drawn based on already existing PAC-Bayesian theorems. In addition, we experimentally evaluate the relevance of CB-Boost in terms of the three main properties we expect about it: accuracy, sparsity, and computational efficiency compared to MinCq, CqBoost, Adaboost and other ensemble methods. As observed in these experiments, CB-Boost not only achieves results comparable to the state of the art, but also provides $$\mathcal {C}$$ -bound sub-optimal weights with very few computational demand while keeping the sparsity property of CqBoost.

We consider the kernel completion problem with the presence of multiple views in the data. In this context the data samples can be fully missing in some views, creating missing columns and rows to the kernel matrices that are calculated individually for each view. We propose to solve the problem of completing the kernel matrices by transferring the features of the other views to represent the view under consideration. We align the known part of the kernel matrix with a new kernel built from the features of the other views. We are thus able to find generalizable structures in the kernel under completion, and represent it accurately. Its missing values can be predicted with the data available in other views. We illustrate the benefits of our approach with simulated data and multivariate digits dataset, as well as with real biological datasets from studies of pattern formation in early \textit{Drosophila melanogaster} embryogenesis.

MOTIVATION Various bioinformatics analyses provide sets of genomic coordinates of interest. Whether two such sets possess a functional relation is a frequent question. This is often determined by interpreting the statistical significance of their overlaps. However, only few existing methods consider the lengths of the overlap, and they do not provide a resolutive p-value. RESULTS Here, we introduce OLOGRAM, which performs overlap statistics between sets of genomic regions described in BEDs or GTF. It uses Monte Carlo simulation, taking into account both the distributions of region and inter-region lengths, to fit a negative binomial model of the total overlap length. Exclusion of user-defined genomic areas during the shuffling is supported. AVAILABILITY This tool is available through the command line interface of the pygtftk toolkit. It has been tested on Linux and OSX and is available on Bioconda and from https://github.com/dputhier/pygtftk under the GNU GPL license. SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

In this article we tackle the supervised multi-view learning problem with kernel methods and metric learning. In this context we consider a recently developed multi-view metric learning (MVML) framework, and propose a SVM-based algorithm that jointly learns the classifier and metrics between views. These metrics permit taking into account the multi-view characteristics of the learning problem. Experiments on real data were performed to evaluate the performance of the proposed algorithm.

We consider the problem of metric learning for multi-view data and present a novel method for learning within-view as well as between-view metrics in vector-valued kernel spaces, as a way to capture multi-modal structure of the data. We formulate two convex optimization problems to jointly learn the metric and the classifier or regressor in kernel feature spaces. An iterative three-step multi-view metric learning algorithm is derived from the optimization problems. In order to scale the computation to large training sets, a blockwise Nystrom approximation of the multi-view kernel matrix is introduced. We justify our approach theoretically and experimentally, and show its performance on real-world datasets against relevant state-of-the-art methods.