We consider learning to optimize a classification metric defined by a black-box function of the confusion matrix. Such black-box learning settings are ubiquitous, for example, when the learner only has query access to the metric of interest, or in noisy-label and domain adaptation applications where the learner must evaluate the metric via performance evaluation using a small validation sample. Our approach is to adaptively learn example weights on the training dataset such that the resulting weighted objective best approximates the metric on the validation sample. We show how to model and estimate the example weights and use them to iteratively post-shift a pre-trained class probability estimator to construct a classifier. We also analyze the resulting procedure’s statistical properties. Experiments on various label noise, domain shift, and fair classification setups confirm that our proposal compares favorably to the state-of-the-art baselines for each application.

We address the problem of training models with black-box and hard-to-optimize metrics by expressing the metric as a monotonic function of a small number of easy-to-optimize surrogates. We pose the training problem as an optimization over a relaxed surrogate space, which we solve by estimating local gradients for the metric and performing inexact convex projections. We analyze gradient estimates based on finite differences and local linear interpolations, and show convergence of our approach under smoothness assumptions with respect to the surrogates. Experimental results on classification and ranking problems verify the proposal performs on par with methods that know the mathematical formulation, and adds notable value when the form of the metric is unknown.

Real-world machine learning applications often have complex test metrics, and may have training and test data that are not identically distributed. Motivated by known connections between complex test metrics and cost-weighted learning, we propose addressing these issues by using a weighted loss function with a standard loss, where the weights on the training examples are learned to optimize the test metric on a validation set. These metric-optimized example weights can be learned for any test metric, including black box and customized ones for specific applications. We illustrate the performance of the proposed method on diverse public benchmark datasets and real-world applications. We also provide a generalization bound for the method.

We introduce the problem of grouping a finite ground set into blocks where each block is a subset of the ground set and where: (i) the blocks are individually highly valued by a submodular function (both robustly and in the average case) while satisfying block-specific matroid constraints; and (ii) block scores interact where blocks are jointly scored highly, thus making the blocks mutually non-redundant. Submodular functions are good models of information and diversity; thus, the above can be seen as grouping the ground set into matroid constrained blocks that are both intra- and inter-diverse. Potential applications include forming ensembles of classification/regression models, partitioning data for parallel processing, and summarization. In the non-robust case, we reduce the problem to non-monotone submodular maximization subject to multiple matroid constraints. In the mixed robust/average case, we offer a bi-criterion guarantee for a polynomial time deterministic algorithm and a probabilistic guarantee for randomized algorithm, as long as the involved submodular functions (including the inter-block interaction terms) are monotone. We close with a case study in which we use these algorithms to find high quality diverse ensembles of classifiers, showing good results.

Practical applications of machine learning often involve successive training iterations with changes to features and training examples. Ideally, changes in the output of any new model should only be improvements (wins) over the previous iteration, but in practice the predictions may change neutrally for many examples, resulting in extra net-zero wins and losses, referred to as unnecessary churn. These changes in the predictions are problematic for usability for some applications, and make it harder and more expensive to measure if a change is statistically significant positive. In this paper, we formulate the problem and present a stabilization operator to regularize a classifier towards a previous classifier. We use a Markov chain Monte Carlo stabilization operator to produce a model with more consistent predictions without adversely affecting accuracy. We investigate the properties of the proposal with theoretical analysis. Experiments on benchmark datasets for different classification algorithms demonstrate the method and the resulting reduction in churn.

For many machine learning problems, there are some inputs that are known to be positively (or negatively) related to the output, and in such cases training the model to respect that monotonic relationship can provide regularization, and makes the model more interpretable. However, flexible monotonic functions are computationally challenging to learn beyond a few features. We break through this barrier by learning ensembles of monotonic calibrated interpolated look-up tables (lattices). A key contribution is an automated algorithm for selecting feature subsets for the ensemble base models. We demonstrate that compared to random forests, these ensembles produce similar or better accuracy, while providing guaranteed monotonicity consistent with prior knowledge, smaller model size and faster evaluation.

Efficiently learning accurate models of dynamical systems is of central importance for developing rational agents that can succeed in a wide range of challenging domains. The difficulty of this learning problem is particularly acute in settings with large observation spaces and partial observability. We present a new algorithm, called Compressed Predictive State Representation (CPSR), for learning models of high-dimensional partially observable uncontrolled dynamical systems from small sample sets. The algorithm, which extends previous work on Predictive State Representations, exploits a particular sparse structure present in many domains. This sparse structure is used to compress information during learning, allowing for an increase in both the efficiency and predictive power. The compression technique also relieves the burden of domain specific feature selection and allows for domains with extremely large discrete observation spaces to be efficiently modelled. We present empirical results showing that the algorithm is able to build accurate models more efficiently than its uncompressed counterparts, and provide theoretical results on the accuracy of the learned compressed model.

This paper addresses the problem of automatic generation of features for value function approximation in reinforcement learning. Bellman Error Basis Functions (BEBFs) have been shown to improve the error of policy evaluation with function approximation, with a convergence rate similar to that of value iteration. We propose a simple, fast and robust algorithm based on random projections, which generates BEBFs for sparse feature spaces. We provide a finite sample analysis of the proposed method, and prove that projections logarithmic in the dimension of the original space guarantee a contraction in the error. Empirical results demonstrate the strength of this method in domains in which choosing a good state representation is challenging.

Recent advances in the area of compressed sensing suggest that it is possible to reconstruct high-dimensional sparse signals from a small number of random projections. Domains in which the sparsity assumption is applicable also offer many interesting large-scale machine learning prediction tasks. It is therefore important to study the effect of random projections as a dimensionality reduction method under such sparsity assumptions. In this paper we develop the bias-variance analysis of a least-squares regression estimator in compressed spaces when random projections are applied on sparse input signals. Leveraging the sparsity assumption, we are able to work with arbitrary non i.i.d. sampling strategies and derive a worst-case bound on the entire space. Empirical results on synthetic and real-world datasets shows how the choice of the projection size affects the performance of regression on compressed spaces, and highlights a range of problems where the method is useful.

Markovian processes have long been used to model stochastic environments. Reinforcement learning has emerged as a framework to solve sequential planning and decision-making problems in such environments. In recent years, attempts were made to apply methods from reinforcement learning to construct decision support systems for action selection in Markovian environments. Although conventional methods in reinforcement learning have proved to be useful in problems concerning sequential decision-making, they cannot be applied in their current form to decision support systems, such as those in medical domains, as they suggest policies that are often highly prescriptive and leave little room for the user's input. Without the ability to provide flexible guidelines, it is unlikely that these methods can gain ground with users of such systems. This paper introduces the new concept of non-deterministic policies to allow more flexibility in the user's decision-making process, while constraining decisions to remain near optimal solutions. We provide two algorithms to compute non-deterministic policies in discrete domains. We study the output and running time of these method on a set of synthetic and real-world problems. In an experiment with human subjects, we show that humans assisted by hints based on non-deterministic policies outperform both human-only and computer-only agents in a web navigation task.

This paper introduces the first set of PAC-Bayesian bounds for the batch reinforcement learning problem in finite state spaces. These bounds hold regardless of the correctness of the prior distribution. We demonstrate how such bounds can be used for model-selection in control problems where prior information is available either on the dynamics of the environment, or on the value of actions. Our empirical results confirm that PAC-Bayesian model-selection is able to leverage prior distributions when they are informative and, unlike standard Bayesian RL approaches, ignores them when they are misleading.

Markov Decision Processes (MDPs) have been extensively studied and used in the context of planning and decision-making, and many methods exist to find the optimal policy for problems modelled as MDPs. Although finding the optimal policy is sufficient in many domains, in certain applications such as decision support systems where the policy is executed by a human (rather than a machine), finding all possible near-optimal policies might be useful as it provides more flexibility to the person executing the policy. In this paper we introduce the new concept of non-deterministic MDP policies, and address the question of finding near-optimal non-deterministic policies. We propose two solutions to this problem, one based on a Mixed Integer Program and the other one based on a search algorithm. We include experimental results obtained from applying this framework to optimize treatment choices in the context of a medical decision support system.