Among all data augmentation techniques proposed so far, linear interpolation of training samples, also called Mixup, has found to be effective for a large panel of applications. Along with improved predictive performance, Mixup is also a good technique for improving calibration. However, mixing data carelessly can lead to manifold mismatch, i.e., synthetic data lying outside original class manifolds, which can deteriorate calibration. In this work, we show that the likelihood of assigning a wrong label with mixup increases with the distance between data to mix. To this end, we propose to dynamically change the underlying distributions of interpolation coefficients depending on the similarity between samples to mix, and define a flexible framework to do so without losing in diversity. We provide extensive experiments for classification and regression tasks, showing that our proposed method improves predictive performance and calibration of models, while being much more efficient.

Developing inherently interpretable models for prediction has gained prominence in recent years. A subclass of these models, wherein the interpretable network relies on learning high-level concepts, are valued because of closeness of concept representations to human communication. However, the visualization and understanding of the learnt unsupervised dictionary of concepts encounters major limitations, especially for large-scale images. We propose here a novel method that relies on mapping the concept features to the latent space of a pretrained generative model. The use of a generative model enables high quality visualization, and lays out an intuitive and interactive procedure for better interpretation of the learnt concepts by imputing concept activations and visualizing generated modifications. Furthermore, leveraging pretrained generative models has the additional advantage of making the training of the system more efficient. We quantitatively ascertain the efficacy of our method in terms of accuracy of the interpretable prediction network, fidelity of reconstruction, as well as faithfulness and consistency of learnt concepts. The experiments are conducted on multiple image recognition benchmarks for large-scale images. Project page available at https://jayneelparekh.github.io/VisCoIN_project_page/

Although graph-based learning has attracted a lot of attention, graph representation learning is still a challenging task whose resolution may impact key application fields such as chemistry or biology. To this end, we introduce GRALE, a novel graph autoencoder that encodes and decodes graphs of varying sizes into a shared embedding space. GRALE is trained using an Optimal Transport-inspired loss that compares the source and reconstructed graphs and leverages a differentiable matching module, which is trained jointly with the encoder and decoder. The proposed attention-based architecture relies on Evoformer, the core component of AlphaFold, which we extend to support both graph encoding and decoding. We show, in numerical experiments on simulated and molecular data, that GRALE enables a highly general form of pre-training, applicable to a wide range of downstream tasks, from classification and regression to more complex tasks such as graph interpolation, editing, matching, and prediction.

Leveraging the kernel trick in both the input and output spaces, surrogate kernel methods are a flexible and theoretically grounded solution to structured output prediction. If they provide state-of-the-art performance on complex data sets of moderate size (e.g., in chemoinformatics), these approaches however fail to scale. We propose to equip surrogate kernel methods with sketching-based approximations, applied to both the input and output feature maps. We prove excess risk bounds on the original structured prediction problem, showing how to attain close-to-optimal rates with a reduced sketch size that depends on the eigendecay of the input/output covariance operators. From a computational perspective, we show that the two approximations have distinct but complementary impacts: sketching the input kernel mostly reduces training time, while sketching the output kernel decreases the inference time. Empirically, our approach is shown to scale, achieving state-of-the-art performance on benchmark data sets where non-sketched methods are intractable.

We propose Any2graph, a generic framework for end-to-end Supervised Graph Prediction (SGP) i.e. a deep learning model that predicts an entire graph for any kind of input. The framework is built on a novel Optimal Transport loss, the Partially-Masked Fused Gromov-Wasserstein, that exhibits all necessary properties (permutation invariance, differentiability and scalability) and is designed to handle any-sized graphs. Numerical experiments showcase the versatility of the approach that outperform existing competitors on a novel challenging synthetic dataset and a variety of real-world tasks such as map construction from satellite image (Sat2Graph) or molecule prediction from fingerprint (Fingerprint2Graph).

This paper tackles two major problem settings for interpretability of audio processing networks,post-hocandby-designinterpretation. For post-hoc interpretation, we aim to interpret decisions of a network in terms of high-level audio objects that are also listenable for the end-user. This is extended to present an inherently interpretable model with high performance. To this end, we propose a novel interpreter design that incorporates non-negative matrix factorization (NMF). In particular, an interpreter is trained to generate a regularized intermediate embedding from hidden layers of a target network, learnt as time-activations of a pre-learnt NMF dictionary. Our methodology allows us to generate intuitive audio-based interpretations that explicitly enhance parts of the input signal most relevant for a network's decision. We demonstrate our method's applicability on a variety of classification tasks, including multi-label data for real-world audio and music.

This paper introduces a novel and generic framework to solve the flagship task of supervised labeled graph prediction by leveraging Optimal Transport tools. We formulate the problem as regression with the Fused Gromov-Wasserstein (FGW) loss and propose a predictive model relying on a FGW barycenter whose weights depend on inputs. First we introduce a non-parametric estimator based on kernel ridge regression for which theoretical results such as consistency and excess risk bound are proved. Next we propose an interpretable parametric model where the barycenter weights are modeled with a neural network and the graphs on which the FGW barycenter is calculated are additionally learned. Numerical experiments show the strength of the method and its ability to interpolate in the labeled graph space on simulated data and on a difficult metabolic identification problem where it can reach very good performance with very little engineering.

The focus of the paper is functional output regression (FOR) with convoluted losses. While most existing work consider the square loss setting, we leverage extensions of the Huber and the $\epsilon$-insensitive loss (induced by infimal convolution) and propose a flexible framework capable of handling various forms of outliers and sparsity in the FOR family. We derive computationally tractable algorithms relying on duality to tackle the resulting tasks in the context of vector-valued reproducing kernel Hilbert spaces. The efficiency of the approach is demonstrated and contrasted with the classical squared loss setting on both synthetic and real-world benchmarks.

This paper tackles post-hoc interpretability for audio processing networks. Our goal is to interpret decisions of a trained network in terms of high-level audio objects that are also listenable for the end-user. To this end, we propose a novel interpreter design that incorporates non-negative matrix factorization (NMF). In particular, a regularized interpreter module is trained to take hidden layer representations of the targeted network as input and produce time activations of pre-learnt NMF components as intermediate outputs. Our methodology allows us to generate intuitive audio-based interpretations that explicitly enhance parts of the input signal most relevant for a network's decision. We demonstrate our method's applicability on popular benchmarks, including a real-world multi-label classification task.

We propose and analyse a reduced-rank method for solving least-squares regression problems with infinite dimensional output. We derive learning bounds for our method, and study under which setting statistical performance is improved in comparison to full-rank method. Our analysis extends the interest of reduced-rank regression beyond the standard low-rank setting to more general output regularity assumptions. We illustrate our theoretical insights on synthetic least-squares problems. Then, we propose a surrogate structured prediction method derived from this reduced-rank method. We assess its benefits on three different problems: image reconstruction, multi-label classification, and metabolite identification.

To address functional-output regression, we introduce projection learning (PL), a novel dictionary-based approach that learns to predict a function that is expanded on a dictionary while minimizing an empirical risk based on a functional loss. PL makes it possible to use non orthogonal dictionaries and can then be combined with dictionary learning; it is thus much more flexible than expansion-based approaches relying on vectorial losses. This general method is instantiated with reproducing kernel Hilbert spaces of vector-valued functions as kernel-based projection learning (KPL). For the functional square loss, two closed-form estimators are proposed, one for fully observed output functions and the other for partially observed ones. Both are backed theoretically by an excess risk analysis. Then, in the more general setting of integral losses based on differentiable ground losses, KPL is implemented using first-order optimization for both fully and partially observed output functions. Eventually, several robustness aspects of the proposed algorithms are highlighted on a toy dataset; and a study on two real datasets shows that they are competitive compared to other nonlinear approaches. Notably, using the square loss and a learnt dictionary, KPL enjoys a particularily attractive trade-off between computational cost and performances.

Originated from Optimal Transport, the Wasserstein distance has gained importance in Machine Learning due to its appealing geometrical properties and the increasing availability of efficient approximations. It owes its recent ubiquity in generative modelling and variational inference to its ability to cope with distributions having non overlapping support. In this work, we consider the problem of estimating the Wasserstein distance between two probability distributions when observations are polluted by outliers. To that end, we investigate how to leverage a Medians of Means (MoM) approach to provide robust estimates. Exploiting the dual Kantorovitch formulation of the Wasserstein distance, we introduce and discuss novel MoM-based robust estimators whose consistency is studied under a data contamination model and for which convergence rates are provided. Beyond computational issues, the choice of the partition size, i.e., the unique parameter of theses robust estimators, is investigated in numerical experiments. Furthermore, these MoM estimators make Wasserstein Generative Adversarial Network (WGAN) robust to outliers, as witnessed by an empirical study on two benchmarks CIFAR10 and Fashion MNIST.

To tackle interpretability in deep learning, we present a novel framework to jointly learn a predictive model and its associated interpretation model. The interpreter provides both local and global interpretability about the predictive model in terms of human-understandable high level attribute functions, with minimal loss of accuracy. This is achieved by a dedicated architecture and well chosen regularization penalties. We seek for a small-size dictionary of high level attribute functions that take as inputs the outputs of selected hidden layers and whose outputs feed a linear classifier. We impose strong conciseness on the activation of attributes with an entropy-based criterion while enforcing fidelity to both inputs and outputs of the predictive model. A detailed pipeline to visualize the learnt features is also developed. Moreover, besides generating interpretable models by design, our approach can be specialized to provide post-hoc interpretations for a pre-trained neural network. We validate our approach against several state-of-the-art methods on multiple datasets and show its efficacy on both kinds of tasks.

One of the challenges in machine learning research is to ensure that presented and published results are sound and reliable. Reproducibility, that is obtaining similar results as presented in a paper or talk, using the same code and data (when available), is a necessary step to verify the reliability of research findings. Reproducibility is also an important step to promote open and accessible research, thereby allowing the scientific community to quickly integrate new findings and convert ideas to practice. Reproducibility also promotes the use of robust experimental workflows, which potentially reduce unintentional errors. In 2019, the Neural Information Processing Systems (NeurIPS) conference, the premier international conference for research in machine learning, introduced a reproducibility program, designed to improve the standards across the community for how we conduct, communicate, and evaluate machine learning research. The program contained three components: a code submission policy, a community-wide reproducibility challenge, and the inclusion of the Machine Learning Reproducibility checklist as part of the paper submission process. In this paper, we describe each of these components, how it was deployed, as well as what we were able to learn from this initiative.

We present a novel automatic system for performing explicit content detection directly on the audio signal. Our modular approach uses an audio-to-character recognition model, a keyword spotting model associated with a dictionary of carefully chosen keywords, and a Random Forest classification model for the final decision. To the best of our knowledge, this is the first explicit content detection system based on audio only. We demonstrate the individual relevance of our modules on a set of sub-tasks and compare our approach to a lyrics-informed oracle and an end-to-end naive architecture. The results obtained are encouraging with a F1-score of 67% on a industrial scale explicit content dataset.

Operator-Valued Kernels (OVKs) and associated vector-valued Reproducing Kernel Hilbert Spaces provide an elegant way to extend scalar kernel methods when the output space is a Hilbert space. Although primarily used in finite dimension for problems like multi-task regression, the ability of this framework to deal with infinite dimensional output spaces unlocks many more applications, such as functional regression, structured output prediction, and structured data representation. However, these sophisticated schemes crucially rely on the kernel trick in the output space, so that most of previous works have focused on the square norm loss function, completely neglecting robustness issues that may arise in such surrogate problems. To overcome this limitation, this paper develops a duality approach that allows to solve OVK machines for a wide range of loss functions. The infinite dimensional Lagrange multipliers are handled through a Double Representer Theorem, and algorithms for \epsilon-insensitive losses and the Huber loss are thoroughly detailed. Robustness benefits are emphasized by a theoretical stability analysis, as well as empirical improvements on structured data applications.

For the purpose of monitoring the behavior of complex infrastructures (\textit{e.g.} aircrafts, transport or energy networks), high-rate sensors are deployed to capture multivariate data, generally unlabeled, in quasi continuous-time to detect quickly the occurrence of anomalies that may jeopardize the smooth operation of the system of interest. The statistical analysis of such massive data of functional nature raises many challenging methodological questions. The primary goal of this paper is to extend the popular {\scshape Isolation Forest} (IF) approach to Anomaly Detection, originally dedicated to finite dimensional observations, to functional data. The major difficulty lies in the wide variety of topological structures that may equip a space of functions and the great variety of patterns that may characterize abnormal curves. We address the issue of (randomly) splitting the functional space in a flexible manner in order to isolate progressively any trajectory from the others, a key ingredient to the efficiency of the algorithm. Beyond a detailed description of the algorithm, computational complexity and stability issues are investigated at length. From the scoring function measuring the degree of abnormality of an observation provided by the proposed variant of the IF algorithm, a \textit{Functional Statistical Depth} function is defined and discussed, as well as a multivariate functional extension. Numerical experiments provide strong empirical evidence of the accuracy of the extension proposed.

Machine learning has witnessed tremendous success in solving tasks depending on a single hyperparameter. When considering simultaneously a finite number of tasks, multi-task learning enables one to account for the similarities of the tasks via appropriate regularizers. A step further consists of learning a continuum of tasks for various loss functions. A promising approach, called Parametric Task Learning, has paved the way in the continuum setting for affine models and piecewise-linear loss functions. In this work, we introduce a novel approach called Infinite Task Learning whose goal is to learn a function whose output is a function over the hyperparameter space. We leverage tools from operator-valued kernels and the associated vector-valued RKHSs that provide an explicit control over the role of the hyperparameters, and also allows us to consider new type of constraints. We provide generalization guarantees to the suggested scheme and illustrate its efficiency in cost-sensitive classification, quantile regression and density level set estimation.

This paper investigates a novel algorithmic approach to data representation based on kernel methods. Assuming that the observations lie in a Hilbert space X , the introduced Kernel Autoencoder (KAE) is the composition of mappings from vector-valued Reproducing Kernel Hilbert Spaces (vv-RKHSs) that minimizes the expected reconstruction error. Beyond a first extension of the autoencoding scheme to possibly infinite dimensional Hilbert spaces, KAE further allows to autoencode any kind of data by choosing X to be itself a RKHS. A theoretical analysis of the model is carried out, providing a generalization bound, and shedding light on its connection with Kernel Principal Component Analysis. The proposed algorithms are then detailed at length: they crucially rely on the form taken by the minimizers, revealed by a dedicated Representer Theorem. Finally, numerical experiments on both simulated data and real labeled graphs (molecules) provide empirical evidence of the KAE performances.

Alexandre Garcia, Pierre Colombo, Florence d’Alché-Buc, Slim Essid, Chloé Clavel. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

Motivated by Supervised Opinion Analysis, we propose a novel framework devoted to Structured Output Learning with Abstention (SOLA). The structure prediction model is able to abstain from predicting some labels in the structured output at a cost chosen by the user in a flexible way. For that purpose, we decompose the problem into the learning of a pair of predictors, one devoted to structured abstention and the other, to structured output prediction. To compare fully labeled training data with predictions potentially containing abstentions, we define a wide class of asymmetric abstention-aware losses. Learning is achieved by surrogate regression in an appropriate feature space while prediction with abstention is performed by solving a new pre-image problem. Thus, SOLA extends recent ideas about Structured Output Prediction via surrogate problems and calibration theory and enjoys statistical guarantees on the resulting excess risk. Instantiated on a hierarchical abstention-aware loss, SOLA is shown to be relevant for fine-grained opinion mining and gives state-of-the-art results on this task. Moreover, the abstention-aware representations can be used to competitively predict user-review ratings based on a sentence-level opinion predictor.

We propose to solve a label ranking problem as a structured output regression task. In this view, we adopt a least square surrogate loss approach that solves a supervised learning problem in two steps: a regression step in a well-chosen feature space and a pre-image (or decoding) step. We use specific feature maps/embeddings for ranking data, which convert any ranking/permutation into a vector representation. These embeddings are all well-tailored for our approach, either by resulting in consistent estimators, or by solving trivially the pre-image problem which is often the bottleneck in structured prediction. Their extension to the case of incomplete or partial rankings is also discussed. Finally, we provide empirical results on synthetic and real-world datasets showing the relevance of our method.

Leveraging the celebrated support vector regression (SVR) method, we propose a unifying framework in order to deliver regression machines in reproducing kernel Hilbert spaces (RKHSs) with data sparsity. The central point is a new definition of $ε$-insensitivity, valid for many regression losses (including quantile and expectile regression) and their multivariate extensions. We show that the dual optimization problem to empirical risk minimization with $ε$-insensitivity involves a data sparse regularization. We also provide an analysis of the excess of risk as well as a randomized coordinate descent algorithm for solving the dual. Numerical experiments validate our approach.

Devoted to multi-task learning and structured output learning, operator-valued kernels provide a flexible tool to build vector-valued functions in the context of Reproducing Kernel Hilbert Spaces. To scale up these methods, we extend the celebrated Random Fourier Feature methodology to get an approximation of operator-valued kernels. We propose a general principle for Operator-valued Random Fourier Feature construction relying on a generalization of Bochner’s theorem for translation-invariant operator-valued Mercer kernels. We prove the uniform convergence of the kernel approximation for bounded and unbounded operator random Fourier features using appropriate Bernstein matrix concentration inequality. An experimental proof-of-concept shows the quality of the approximation and the efficiency of the corresponding linear models on example datasets.

Addressing the will to give a more complete picture than an average relationship provided by standard regression, a novel framework for estimating and predicting simultaneously several conditional quantiles is introduced. The proposed methodology leverages kernel-based multi-task learning to curb the embarrassing phenomenon of quantile crossing, with a one-step estimation procedure and no post-processing. Moreover, this framework comes along with theoretical guarantees and an efficient coordinate descent learning algorithm. Numerical experiments on benchmark and real datasets highlight the enhancements of our approach regarding the prediction error, the crossing occurrences and the training time.

MOTIVATION: An important problematic of metabolomics is to identify metabolites using tandem mass spectrometry data. Machine learning methods have been proposed recently to solve this problem by predicting molecular fingerprint vectors and matching these fingerprints against existing molecular structure databases. In this work we propose to address the metabolite identification problem using a structured output prediction approach. This type of approach is not limited to vector output space and can handle structured output space such as the molecule space. RESULTS: We use the Input Output Kernel Regression method to learn the mapping between tandem mass spectra and molecular structures. The principle of this method is to encode the similarities in the input (spectra) space and the similarities in the output (molecule) space using two kernel functions. This method approximates the spectra-molecule mapping in two phases. The first phase corresponds to a regression problem from the input space to the feature space associated to the output kernel. The second phase is a preimage problem, consisting in mapping back the predicted output feature vectors to the molecule space. We show that our approach achieves state-of-the-art accuracy in metabolite identification. Moreover, our method has the advantage of decreasing the running times for the training step and the test step by several orders of magnitude over the preceding methods. CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

In this paper, we introduce a novel approach, called Input Output Kernel Regression (IOKR), for learning mappings between structured inputs and structured outputs. The approach belongs to the family of Output Kernel Regression methods devoted to regression in feature space endowed with some output kernel. In order to take into account structure in input data and benefit from kernels in the input space as well, we use the Reproducing Kernel Hilbert Space theory for vector-valued functions. We first recall the ridge solution for supervised learning and then study the regularized hinge loss-based solution used in Maximum Margin Regression. Both models are also developed in the context of semi-supervised setting. In addition we derive an extension of Generalized Cross Validation for model selection in the case of the least-square model. Finally we show the versatility of the IOKR framework on two different problems: link prediction seen as a structured output problem and multi-task regression seen as a multiple and interdependent output problem. Eventually, we present a set of detailed numerical results that shows the relevance of the method on these two tasks.

MOTIVATION: Identifying the set of genes differentially expressed along time is an important task in two-sample time course experiments. Furthermore, estimating at which time periods the differential expression is present can provide additional insight into temporal gene functions. The current differential detection methods are designed to detect difference along observation time intervals or on single measurement points, warranting dense measurements along time to characterize the full temporal differential expression patterns. RESULTS: We propose a novel Bayesian likelihood ratio test to estimate the differential expression time periods. Applying the ratio test to systems of genes provides the temporal response timings and durations of gene expression to a biological condition. We introduce a novel non-stationary Gaussian process as the underlying expression model, with major improvements on model fitness on perturbation and stress experiments. The method is robust to uneven or sparse measurements along time. We assess the performance of the method on realistically simulated dataset and compare against state-of-the-art methods. We additionally apply the method to the analysis of primary human endothelial cells under an ionizing radiation stress to study the transcriptional perturbations over 283 measured genes in an attempt to better understand the role of endothelium in both normal and cancer tissues during radiotherapy. As a result, using the cascade of differential expression periods, domain literature and gene enrichment analysis, we gain insights into the dynamic response of endothelial cells to irradiation. AVAILABILITY AND IMPLEMENTATION: R package 'nsgp' is available at www.ibisc.fr/en/logiciels_arobas.

Computational methods for predicting protein-protein interactions are important tools that can complement high-throughput technologies and guide biologists in designing new laboratory experiments. The proteins and the interactions between them can be described by a network which is characterized by several topological properties. Information about proteins and interactions between them, in combination with knowledge about topological properties of the network, can be used for developing computational methods that can accurately predict unknown protein-protein interactions. This paper presents a supervised learning framework based on Bayesian inference for combining two types of information: i) network topology information, and ii) information related to proteins and the interactions between them. The motivation of our model is that by combining these two types of information one can achieve a better accuracy in predicting protein-protein interactions, than by using models constructed from these two types of information independently.

MOTIVATION: Reverse engineering of gene regulatory networks remains a central challenge in computational systems biology, despite recent advances facilitated by benchmark in silico challenges that have aided in calibrating their performance. A number of approaches using either perturbation (knock-out) or wild-type time-series data have appeared in the literature addressing this problem, with the latter using linear temporal models. Nonlinear dynamical models are particularly appropriate for this inference task, given the generation mechanism of the time-series data. In this study, we introduce a novel nonlinear autoregressive model based on operator-valued kernels that simultaneously learns the model parameters, as well as the network structure. RESULTS: A flexible boosting algorithm (OKVAR-Boost) that shares features from L2-boosting and randomization-based algorithms is developed to perform the tasks of parameter learning and network inference for the proposed model. Specifically, at each boosting iteration, a regularized Operator-valued Kernel-based Vector AutoRegressive model (OKVAR) is trained on a random subnetwork. The final model consists of an ensemble of such models. The empirical estimation of the ensemble model's Jacobian matrix provides an estimation of the network structure. The performance of the proposed algorithm is first evaluated on a number of benchmark datasets from the DREAM3 challenge and then on real datasets related to the In vivo Reverse-Engineering and Modeling Assessment (IRMA) and T-cell networks. The high-quality results obtained strongly indicate that it outperforms existing approaches. AVAILABILITY: The OKVAR-Boost Matlab code is available as the archive: http://amis-group.fr/sourcecode-okvar-boost/OKVARBoost-v1.0.zip. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Gene regulatory network inference remains a challenging problem in systems biology despite the numerous approaches that have been proposed. When substantial knowledge on a gene regulatory network is already available, supervised network inference is appropriate. Such a method builds a binary classifier able to assign a class (Regulation/No regulation) to an ordered pair of genes. Once learnt, the pairwise classifier can be used to predict new regulations. In this work, we explore the framework of Markov Logic Networks (MLN) that combine features of probabilistic graphical models with the expressivity of first-order logic rules. RESULTS: We propose to learn a Markov Logic network, e.g. a set of weighted rules that conclude on the predicate "regulates", starting from a known gene regulatory network involved in the switch proliferation/differentiation of keratinocyte cells, a set of experimental transcriptomic data and various descriptions of genes all encoded into first-order logic. As training data are unbalanced, we use asymmetric bagging to learn a set of MLNs. The prediction of a new regulation can then be obtained by averaging predictions of individual MLNs. As a side contribution, we propose three in silico tests to assess the performance of any pairwise classifier in various network inference tasks on real datasets. A first test consists of measuring the average performance on balanced edge prediction problem; a second one deals with the ability of the classifier, once enhanced by asymmetric bagging, to update a given network. Finally our main result concerns a third test that measures the ability of the method to predict regulations with a new set of genes. As expected, MLN, when provided with only numerical discretized gene expression data, does not perform as well as a pairwise SVM in terms of AUPR. However, when a more complete description of gene properties is provided by heterogeneous sources, MLN achieves the same performance as a black-box model such as a pairwise SVM while providing relevant insights on the predictions. CONCLUSIONS: The numerical studies show that MLN achieves very good predictive performance while opening the door to some interpretability of the decisions. Besides the ability to suggest new regulations, such an approach allows to cross-validate experimental data with existing knowledge.

SUMMARY: CycSim is a web application dedicated to in silico experiments with genome-scale metabolic models coupled to the exploration of knowledge from BioCyc and KEGG. Specifically, CycSim supports the design of knockout experiments: simulation of growth phenotypes of single or multiple gene deletions mutants on specified media, comparison of these predictions with experimental phenotypes and direct visualization of both on metabolic maps. The web interface is designed for simplicity, putting constraint-based modelling techniques within easier reach of biologists. CycSim also functions as an online repository of genome-scale metabolic models. AVAILABILITY: http://www.genoscope.cns.fr/cycsim.

BACKGROUND: Inferring gene regulatory networks from data requires the development of algorithms devoted to structure extraction. When only static data are available, gene interactions may be modelled by a Bayesian Network (BN) that represents the presence of direct interactions from regulators to regulees by conditional probability distributions. We used enhanced evolutionary algorithms to stochastically evolve a set of candidate BN structures and found the model that best fits data without prior knowledge. RESULTS: We proposed various evolutionary strategies suitable for the task and tested our choices using simulated data drawn from a given bio-realistic network of 35 nodes, the so-called insulin network, which has been used in the literature for benchmarking. We assessed the inferred models against this reference to obtain statistical performance results. We then compared performances of evolutionary algorithms using two kinds of recombination operators that operate at different scales in the graphs. We introduced a niching strategy that reinforces diversity through the population and avoided trapping of the algorithm in one local minimum in the early steps of learning. We show the limited effect of the mutation operator when niching is applied. Finally, we compared our best evolutionary approach with various well known learning algorithms (MCMC, K2, greedy search, TPDA, MMHC) devoted to BN structure learning. CONCLUSION: We studied the behaviour of an evolutionary approach enhanced by niching for the learning of gene regulatory networks with BN. We show that this approach outperforms classical structure learning methods in elucidating the original model. These results were obtained for the learning of a bio-realistic network and, more importantly, on various small datasets. This is a suitable approach for learning transcriptional regulatory networks from real datasets without prior knowledge.

A general framework is proposed for gradient boosting in supervised learning problems where the loss function is defined using a kernel over the output space. It extends boosting in a principled way to complex output spaces (images, text, graphs etc.) and can be applied to a general class of base learners working in kernelized output spaces. Empirical results are provided on three problems: a regression problem, an image completion task and a graph prediction problem. In these experiments, the framework is combined with tree-based base learners, which have interesting algorithmic properties. The results show that gradient boosting significantly improves these base learners and provides competitive results with other tree-based ensemble methods based on randomization.

MOTIVATION: Statistical inference of biological networks such as gene regulatory networks, signaling pathways and metabolic networks can contribute to build a picture of complex interactions that take place in the cell. However, biological systems considered as dynamical, non-linear and generally partially observed processes may be difficult to estimate even if the structure of interactions is given. RESULTS: Using the same approach as Sitz et al. proposed in another context, we derive non-linear state-space models from ODEs describing biological networks. In this framework, we apply Unscented Kalman Filtering (UKF) to the estimation of both parameters and hidden variables of non-linear state-space models. We instantiate the method on a transcriptional regulatory model based on Hill kinetics and a signaling pathway model based on mass action kinetics. We successfully use synthetic data and experimental data to test our approach. CONCLUSION: This approach covers a large set of biological networks models and gives rise to simple and fast estimation algorithms. Moreover, the Bayesian tool used here directly provides uncertainty estimates on parameters and hidden states. Let us also emphasize that it can be coupled with structure inference methods used in Graphical Probabilistic Models. AVAILABILITY: Matlab code available on demand.

BACKGROUND: Elucidating biological networks between proteins appears nowadays as one of the most important challenges in systems biology. Computational approaches to this problem are important to complement high-throughput technologies and to help biologists in designing new experiments. In this work, we focus on the completion of a biological network from various sources of experimental data. RESULTS: We propose a new machine learning approach for the supervised inference of biological networks, which is based on a kernelization of the output space of regression trees. It inherits several features of tree-based algorithms such as interpretability, robustness to irrelevant variables, and input scalability. We applied this method to the inference of two kinds of networks in the yeast S. cerevisiae: a protein-protein interaction network and an enzyme network. In both cases, we obtained results competitive with existing approaches. We also show that our method provides relevant insights on input data regarding their potential relationship with the existence of interactions. Furthermore, we confirm the biological validity of our predictions in the context of an analysis of gene expression data. CONCLUSION: Output kernel tree based methods provide an efficient tool for the inference of biological networks from experimental data. Their simplicity and interpretability should make them of great value for biologists.

We extend tree-based methods to the prediction of structured outputs using a kernelization of the algorithm that allows one to grow trees as soon as a kernel can be defined on the output space. The resulting algorithm, called output kernel trees (OK3), generalizes classification and regression trees as well as tree-based ensemble methods in a principled way. It inherits several features of these methods such as interpretability, robustness to irrelevant variables, and input scalability. When only the Gram matrix over the outputs of the learning sample is given, it learns the output kernel as a function of inputs. We show that the proposed algorithm works well on an image reconstruction task and on a biological network inference problem.

In this paper, we propose a new model, the kernel Kalman Filter, to perform various nonlinear time series processing. This model is based on the use of Mercer kernel functions in the framework of the Kalman filter or linear dynamical systems. Thanks to the kernel trick, all the equations involved in our model to perform filtering, smoothing and learning tasks, only require matrix algebra calculus whilst providing the ability to model complex time series. In particular, it is possible to learn dynamics from some nonlinear noisy time series implementing an exact expectation-maximization procedure.

We consider the question of predicting nonlinear time series. Kernel Dy- namical Modeling (KDM), a new method based on kernels, is proposed as an extension to linear dynamical models. The kernel trick is used twice: first, to learn the parameters of the model, and second, to compute preimages of the time series predicted in the feature space by means of Support Vector Regression. Our model shows strong connection with the classic Kalman Filter model, with the kernel feature space as hidden state space. Kernel Dynamical Modeling is tested against two benchmark time series and achieves high quality predictions.

In many discrimination problems a large amount of data is available but only a few of them are labeled. This provides a strong motivation to improve or develop methods for semi-supervised learning. In this paper, boosting is generalized to this task within the optimization framework of MarginBoost . We extend the margin definition to unlabeled data and develop the gradient descent algorithm that corresponds to the resulting margin cost function. This meta-learning scheme can be applied to any base classifier able to benefit from unlabeled data. We propose here to apply it to mixture models trained with an Expectation-Maximization algorithm. Promising results are presented on benchmarks with different rates of labeled data.

We propose to solve a text categorization task using a new metric between documents, based on a priori semantic knowledge about words. This metric can be incorporated into the definition of radial basis kernels of Support Vector Machines or directly used in a K-nearest neighbors algorithm. Both SVM and KNN are tested and compared on the 20-newsgroups database. Support Vector Machines provide the best accuracy on test data.

This paper deals with the learning of understandable decision rules with connectionist systems. Our approach consists of extracting fuzzy control rules with a new fuzzy neural network. Whereas many other works on this area propose to use combinations of nonlinear neurons to approximate fuzzy operations, we use a fuzzy neuron that computes max-min operations. Thus, this neuron can be interpreted as a possibility estimator, just as sigma-pi neurons can support a probabilistic interpretation. Within this context, possibilistic inferences can be drawn through the multi-layered network, using a distributed representation of the information. A new learning procedure has been developed in order that each part of the network can be learnt sequentially, while other parts are frozen. Each step of the procedure is based on the same kind of learning scheme: the backpropagation of a well-chosen cost function with appropriate derivatives of max-min function. An appealing result of the learning phase is the ability of the network to automatically reduce the number of the condition-parts of the rules, if needed. The network has been successfully tested on the learning of a control rule base for an inverted pendulum.

Neural trees are constructive algorithms which build decision trees whose nodes are binary neurons. We propose a new learning scheme, "trio-learning," which leads to a significant reduction in the tree complexity. In this strategy, each node of the tree is optimized by taking into account the knowledge that it will be followed by two son nodes. Moreover, trio-learning can be used to build hybrid trees, with internal nodes and terminal nodes of different nature, for solving any standard tasks (e.g. classification, regression, density estimation). Significant results on a handwritten character classification are presented.