We propose to use input convex neural networks (ICNN) to build convex approximations of non-convex feasible sets of optimization problems, in the form of a set of linear equalities and inequalities in a lifted space.Our approach may be tailored to yield both inner-and outer-approximations, or to maximize its accuracy in regions closer to the minimum of a given objective function.We illustrate the method on twodimensional toy problems and motivate it by various instances of reliability management problems of large-scale electric power systems.

Random forests have been widely used for their ability to provide so-called importance measures, which give insight at a global (per dataset) level on the relevance of input variables to predict a certain output. On the other hand, methods based on Shapley values have been introduced to refine the analysis of feature relevance in tree-based models to a local (per instance) level. In this context, we first show that the global Mean Decrease of Impurity (MDI) variable importance scores correspond to Shapley values under some conditions. Then, we derive a local MDI importance measure of variable relevance, which has a very natural connection with the global MDI measure and can be related to a new notion of local feature relevance. We further link local MDI importances with Shapley values and discuss them in the light of related measures from the literature. The measures are illustrated through experiments on several classification and regression problems.

This article reviews recent works applying machine learning (ML) techniques in the context of energy systems' reliability assessment and control. We showcase both the progress achieved to date as well as the important future directions for further research, while providing an adequate background in the fields of reliability management and of ML. The objective is to foster the synergy between these two fields and speed up the practical adoption of ML techniques for energy systems reliability management. We focus on bulk electric power systems and use them as an example, but we argue that the methods, tools, etc. can be extended to other similar systems, such as distribution systems, microgrids, and multienergy systems.

Dealing with datasets of very high dimension is a major challenge in machine learning. In this paper, we consider the problem of feature selection in applications where the memory is not large enough to contain all features. In this setting, we propose a novel tree-based feature selection approach that builds a sequence of randomized trees on small subsamples of variables mixing both variables already identified as relevant by previous models and variables randomly selected among the other variables. As our main contribution, we provide an in-depth theoretical analysis of this method in infinite sample setting. In particular, we study its soundness with respect to common definitions of feature relevance and its convergence speed under various variable dependance scenarios. We also provide some preliminary empirical results highlighting the potential of the approach.

We present in this paper a new generic approach to variable branching in branch and bound for mixed-integer linear problems. Our approach consists in imitating the decisions taken by a good branching strategy, namely strong branching, with a fast approximation. This approximated function is created by a machine learning technique from a set of observed branching decisions taken by strong branching. The philosophy of the approach is similar to reliability branching. However, our approach can catch more complex aspects of observed previous branchings to take a branching decision. The experiments performed on randomly generated and MIPLIB problems show promising results.

MOTIVATION: Collaborative analysis of massive imaging datasets is essential to enable scientific discoveries. RESULTS: We developed Cytomine to foster active and distributed collaboration of multidisciplinary teams for large-scale image-based studies. It uses web development methodologies and machine learning in order to readily organize, explore, share and analyze (semantically and quantitatively) multi-gigapixel imaging data over the internet. We illustrate how it has been used in several biomedical applications. AVAILABILITY AND IMPLEMENTATION: Cytomine (http://www.cytomine.be/) is freely available under an open-source license from http://github.com/cytomine/ A documentation wiki (http://doc.cytomine.be) and a demo server (http://demo.cytomine.be) are also available. CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Fine operating rules for security control and an automatic system for their online discovery were developed to adapt to the development of smart grids. The automatic system uses the real-time system state to determine critical flowgates, and then a continuation power flow-based security analysis is used to compute the initial transfer capability of critical flowgates. Next, the system applies the Monte Carlo simulations to expected short-term operating condition changes, feature selection, and a linear least squares fitting of the fine operating rules. The proposed system was validated both on an academic test system and on a provincial power system in China. The results indicated that the derived rules provide accuracy and good interpretability and are suitable for real-time power system security control. The use of high-performance computing systems enables these fine operating rules to be refreshed online every 15 min.

Despite growing interest and practical use in various scientific areas, variable importances derived from tree-based ensemble methods are not well understood from a theoretical point of view. In this work we characterize the Mean Decrease Impurity (MDI) variable importances as measured by an ensemble of totally randomized trees in asymptotic sample and ensemble size conditions. We derive a three-level decomposition of the information jointly provided by all input variables about the output in terms of i) the MDI importance of each input variable, ii) the degree of interaction of a given input variable with the other input variables, iii) the different interaction terms of a given degree. We then show that this MDI importance of a variable is equal to zero if and only if the variable is irrelevant and that the MDI importance of a relevant variable is invariant with respect to the removal or the addition of irrelevant variables. We illustrate these properties on a simple example and discuss how they may change in the case of non-totally randomized trees such as Random Forests and Extra-Trees.

BACKGROUND: Research in epistasis or gene-gene interaction detection for human complex traits has grown over the last few years. It has been marked by promising methodological developments, improved translation efforts of statistical epistasis to biological epistasis and attempts to integrate different omics information sources into the epistasis screening to enhance power. The quest for gene-gene interactions poses severe multiple-testing problems. In this context, the maxT algorithm is one technique to control the false-positive rate. However, the memory needed by this algorithm rises linearly with the amount of hypothesis tests. Gene-gene interaction studies will require a memory proportional to the squared number of SNPs. A genome-wide epistasis search would therefore require terabytes of memory. Hence, cache problems are likely to occur, increasing the computation time. In this work we present a new version of maxT, requiring an amount of memory independent from the number of genetic effects to be investigated. This algorithm was implemented in C++ in our epistasis screening software MBMDR-3.0.3. We evaluate the new implementation in terms of memory efficiency and speed using simulated data. The software is illustrated on real-life data for Crohn's disease. RESULTS: In the case of a binary (affected/unaffected) trait, the parallel workflow of MBMDR-3.0.3 analyzes all gene-gene interactions with a dataset of 100,000 SNPs typed on 1000 individuals within 4 days and 9 hours, using 999 permutations of the trait to assess statistical significance, on a cluster composed of 10 blades, containing each four Quad-Core AMD Opteron(tm) Processor 2352 2.1 GHz. In the case of a continuous trait, a similar run takes 9 days. Our program found 14 SNP-SNP interactions with a multiple-testing corrected p-value of less than 0.05 on real-life Crohn's disease (CD) data. CONCLUSIONS: Our software is the first implementation of the MB-MDR methodology able to solve large-scale SNP-SNP interactions problems within a few days, without using much memory, while adequately controlling the type I error rates. A new implementation to reach genome-wide epistasis screening is under construction. In the context of Crohn's disease, MBMDR-3.0.3 could identify epistasis involving regions that are well known in the field and could be explained from a biological point of view. This demonstrates the power of our software to find relevant phenotype-genotype higher-order associations.

In the context of multistage stochastic optimization problems, we propose a hybrid strategy for generalizing to nonlinear decision rules, using machine learning, a finite data set of constrained vector-valued recourse decisions optimized using scenario-tree techniques from multistage stochastic programming. The decision rules are based on a statistical model inferred from a given scenario-tree solution and are selected by out-of-sample simulation given the true problem. Because the learned rules depend on the given scenario tree, we repeat the procedure for a large number of randomly generated scenario trees and then select the best solution (policy) found for the true problem. The scheme leads to an ex post selection of the scenario tree itself. Numerical tests evaluate the dependence of the approach on the machine learning aspects and show cases where one can obtain near-optimal solutions, starting with a “weak” scenario-tree generator that randomizes the branching structure of the trees.

MOTIVATION: Univariate statistical tests are widely used for biomarker discovery in bioinformatics. These procedures are simple, fast and their output is easily interpretable by biologists but they can only identify variables that provide a significant amount of information in isolation from the other variables. As biological processes are expected to involve complex interactions between variables, univariate methods thus potentially miss some informative biomarkers. Variable relevance scores provided by machine learning techniques, however, are potentially able to highlight multivariate interacting effects, but unlike the p-values returned by univariate tests, these relevance scores are usually not statistically interpretable. This lack of interpretability hampers the determination of a relevance threshold for extracting a feature subset from the rankings and also prevents the wide adoption of these methods by practicians. RESULTS: We evaluated several, existing and novel, procedures that extract relevant features from rankings derived from machine learning approaches. These procedures replace the relevance scores with measures that can be interpreted in a statistical way, such as p-values, false discovery rates, or family wise error rates, for which it is easier to determine a significance level. Experiments were performed on several artificial problems as well as on real microarray datasets. Although the methods differ in terms of computing times and the tradeoff, they achieve in terms of false positives and false negatives, some of them greatly help in the extraction of truly relevant biomarkers and should thus be of great practical interest for biologists and physicians. As a side conclusion, our experiments also clearly highlight that using model performance as a criterion for feature selection is often counter-productive. AVAILABILITY AND IMPLEMENTATION: Python source codes of all tested methods, as well as the MATLAB scripts used for data simulation, can be found in the Supplementary Material.

We propose a strategy for experiment selection - in the context of reinforcement learning - based on the idea that the most interesting experiments to carry out at some stage are those that are the most liable to falsify the current hypothesis about the optimal control policy. We cast this idea in a context where a policy learning algorithm and a model identification method are given a priori. Experiments are selected if, using the learnt environment model, they are predicted to yield a revision of the learnt control policy. Algorithms and simulation results are provided for a deterministic system with discrete action space. They show that the proposed approach is promising.

Optimizing decisions on an ensemble of incomplete disturbance trees and aggregating their first stage decisions has been shown as a promising approach to (model-based) planning under uncertainty in large continuous action spaces and in small discrete ones. The present paper extends this approach and deals with large but highly structured action spaces, through a kernel-based aggregation scheme. The technique is applied to a test problem with a discrete action space of 6561 elements adapted from the NIPS 2005 SensorNetwork benchmark.

We propose an approach for inferring bounds on the finite-horizon return of a control policy from an off-policy sample of trajectories collecting state transitions, rewards, and control actions. In this paper, the dynamics, control policy, and reward function are supposed to be deterministic and Lipschitz continuous. Under these assumptions, a polynomial algorithm, in terms of the sample size and length of the optimization horizon, is derived to compute these bounds, and their tightness is characterized in terms of the sample density.

This paper compares reinforcement learning (RL) with model predictive control (MPC) in a unified framework and reports experimental results of their application to the synthesis of a controller for a nonlinear and deterministic electrical power oscillations damping problem. Both families of methods are based on the formulation of the control problem as a discrete-time optimal control problem. The considered MPC approach exploits an analytical model of the system dynamics and cost function and computes open-loop policies by applying an interior-point solver to a minimization problem in which the system dynamics are represented by equality constraints. The considered RL approach infers in a model-free way closed-loop policies from a set of system trajectories and instantaneous cost values by solving a sequence of batch-mode supervised learning problems. The results obtained provide insight into the pros and cons of the two approaches and show that RL may certainly be competitive with MPC even in contexts where a good deterministic system model is available.

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.

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.

We present a novel, generic image classification method based on a recent machine learning algorithm (ensembles of extremely randomized decision trees). Images are classified using randomly extracted subwindows that are suitably normalized to yield robustness to certain image transformations. Our method is evaluated on four very different, publicly available datasets (COIL-100, ZuBuD, ETH-80, WANG). Our results show that our automatic approach is generic and robust to illumination, scale, and viewpoint changes. An extension of the method is proposed to improve its robustness with respect to rotation changes.

This paper studies the aggregation of predictions made by tree-based models for several perturbed versions of the attribute vector of a test case. A closed-form approximation of this scheme combined with cross-validation to tune the level of perturbation is proposed. This yields soft-tree models in a parameter free way. and preserves their interpretability. Empirical evaluations, on classification and regression problems, show that accuracy and bias/variance tradeoff are improved significantly at the price of an acceptable computational overhead. The method is further compared and combined with tree bagging.

Motivation: Modern mass spectrometry allows the determination of proteomic fingerprints of body fluids like serum, saliva or urine. These measurements can be used in many medical applications in order to diagnose the current state or predict the evolution of a disease. Recent developments in machine learning allow one to exploit such datasets, characterized by small numbers of very high-dimensional samples. Results: We propose a systematic approach based on decision tree ensemble methods, which is used to automatically determine proteomic biomarkers and predictive models. The approach is validated on two datasets of surface-enhanced laser desorption/ionization time of flight measurements, for the diagnosis of rheumatoid arthritis and inflammatory bowel diseases. The results suggest that the methodology can handle a broad class of similar problems. Supplementary information: Additional tables, appendicies and datasets may be found at http://www.montefiore.ulg.ac.be/~geurts/Papers/Proteomic-suppl.html Contact: [email protected]

Reinforcement learning aims to determine an optimal control policy from interaction with a system or from observations gathered from a system. In batch mode, it can be achieved by approximating the so-called Q-function based on a set of four-tuples (xt, ut , rt, xt+1) where xt denotes the system state at time t, ut the control action taken, rt the instantaneous reward obtained and xt+1 the successor state of the system, and by determining the control policy from this Q-function. The Q-function approximation may be obtained from the limit of a sequence of (batch mode) supervised learning problems. Within this framework we describe the use of several classical tree-based supervised learning methods (CART, Kd-tree, tree bagging) and two newly proposed ensemble algorithms, namely extremely and totally randomized trees. We study their performances on several examples and find that the ensemble methods based on regression trees perform well in extracting relevant information about the optimal control policy from sets of four-tuples. In particular, the totally randomized trees give good results while ensuring the convergence of the sequence, whereas by relaxing the convergence constraint even better accuracy results are provided by the extremely randomized trees.