Supervised classification techniques use training samples to find classification rules with small expected 0-1 loss. Conventional methods achieve efficient learning and out-of-sample generalization by minimizing surrogate losses over specific families of rules. This paper presents minimax risk classifiers (MRCs) that do not rely on a choice of surrogate loss and family of rules. MRCs achieve efficient learning and out-of-sample generalization by minimizing worst-case expected 0-1 loss w.r.t. uncertainty sets that are defined by linear constraints and include the true underlying distribution. In addition, MRCs' learning stage provides performance guarantees as lower and upper tight bounds for expected 0-1 loss. We also present MRCs' finite-sample generalization bounds in terms of training size and smallest minimax risk, and show their competitive classification performance w.r.t. state-of-the-art techniques using benchmark datasets.

The general problem in this paper is vertex (node) subset selection with the goal to contain an infection that spreads in a network. Instead of selecting the single most important node, this paper deals with the problem of selecting multiple nodes for removal. As compared to previous work on multiple-node selection, the trade-off between cost and benefit is considered. The benefit is measured in terms of increasing the epidemic threshold which is a measure of how difficult it is for an infection to spread in a network. The cost is measured in terms of the number and size of nodes to be removed or controlled. Already in its single-objective instance with a fixed number of $k$ nodes to be removed, the multiple vertex immunisation problems have been proven to be NP-hard. Several heuristics have been developed to approximate the problem. In this work, we compare meta-heuristic techniques with exact methods on the Shield-value, which is a sub-modular proxy for the maximal eigenvalue and used in the current state-of-the-art greedy node-removal strategies. We generalise it to the multi-objective case and replace the greedy algorithm by a quadratic program (QP), which then can be solved with exact QP solvers. The main contribution of this paper is the insight that, if time permits, exact and problem-specific methods approximation should be used, which are often far better than Pareto front approximations obtained by general meta-heuristics. Based on these, it will be more effective to develop strategies for controlling real-world networks when the goal is to prevent or contain epidemic outbreaks. This paper is supported by ready to use Python implementation of the optimization methods and datasets.

We consider the problem of black-box function optimization over the boolean hypercube. Despite the vast literature on black-box function optimization over continuous domains, not much attention has been paid to learning models for optimization over combinatorial domains until recently. However, the computational complexity of the recently devised algorithms are prohibitive even for moderate numbers of variables; drawing one sample using the existing algorithms is more expensive than a function evaluation for many black-box functions of interest. To address this problem, we propose a computationally efficient model learning algorithm based on multilinear polynomials and exponential weight updates. In the proposed algorithm, we alternate between simulated annealing with respect to the current polynomial representation and updating the weights using monomial experts' advice. Numerical experiments on various datasets in both unconstrained and sum-constrained boolean optimization indicate the competitive performance of the proposed algorithm, while improving the computational time up to several orders of magnitude compared to state-of-the-art algorithms in the literature.

We present an algorithm for local, regularized, policy improvement in reinforcement learning (RL) that allows us to formulate model-based and model-free variants in a single framework. Our algorithm can be interpreted as a natural extension of work on KL-regularized RL and introduces a form of tree search for continuous action spaces. We demonstrate that additional computation spent on model-based policy improvement during learning can improve data efficiency, and confirm that model-based policy improvement during action selection can also be beneficial. Quantitatively, our algorithm improves data efficiency on several continuous control benchmarks (when a model is learned in parallel), and it provides significant improvements in wall-clock time in high-dimensional domains (when a ground truth model is available). The unified framework also helps us to better understand the space of model-based and model-free algorithms. In particular, we demonstrate that some benefits attributed to model-based RL can be obtained without a model, simply by utilizing more computation.

In this paper we consider multiple Automated Guided Vehicles (AGVs) navigating a common workspace to fulfill various intralogistics tasks, typically formulated as the Multi-Agent Path Finding (MAPF) problem. To keep plan execution deadlock-free, one approach is to construct an Action Dependency Graph (ADG) which encodes the ordering of AGVs as they proceed along their routes. Using this method, delayed AGVs occasionally require others to wait for them at intersections, thereby affecting the plan execution efficiency. If the workspace is shared by dynamic obstacles such as humans or third party robots, AGVs can experience large delays. A common mitigation approach is to re-solve the MAPF using the current, delayed AGV positions. However, solving the MAPF is time-consuming, making this approach inefficient, especially for large AGV teams. In this work, we present an online method to repeatedly modify a given acyclic ADG to minimize route completion times of each AGV. Our approach persistently maintains an acyclic ADG, necessary for deadlock-free plan execution. We evaluate the approach by considering simulations with random disturbances on the execution and show faster route completion times compared to the baseline ADG-based execution management approach.

Neural architecture search (NAS) has attracted a lot of attention and has been illustrated to bring tangible benefits in a large number of applications in the past few years. Network topology and network size have been regarded as two of the most important aspects for the performance of deep learning models and the community has spawned lots of searching algorithms for both aspects of the neural architectures. However, the performance gain from these searching algorithms is achieved under different search spaces and training setups. This makes the overall performance of the algorithms to some extent incomparable and the improvement from a sub-module of the searching model unclear. In this paper, we propose NATS-Bench, a unified benchmark on searching for both topology and size, for (almost) any up-to-date NAS algorithm. NATS-Bench includes the search space of 15,625 neural cell candidates for architecture topology and 32,768 for architecture size on three datasets. We analyse the validity of our benchmark in terms of various criteria and performance comparison of all candidates in the search space. We also show the versatility of NATS-Bench by benchmarking 13 recent state-of-the-art NAS algorithms on it. All logs and diagnostic information trained using the same setup for each candidate are provided. This facilitates a much larger community of researchers to focus on developing better NAS algorithms in a more comparable and computationally cost friendly environment. All codes are publicly available at: https://xuanyidong.com/assets/projects/NATS-Bench .

Various model-based diagnosis scenarios require the computation of most preferred fault explanations. Existing algorithms that are sound (i.e., output only actual fault explanations) and complete (i.e., can return all explanations), however, require exponential space to achieve this task. As a remedy, we propose two novel diagnostic search algorithms, called RBF-HS (Recursive Best-First Hitting Set Search) and HBF-HS (Hybrid Best-First Hitting Set Search), which build upon tried and tested techniques from the heuristic search domain. RBF-HS can enumerate an arbitrary predefined finite number of fault explanations in best-first order within linear space bounds, without sacrificing the desirable soundness or completeness properties. The idea of HBF-HS is to find a trade-off between runtime optimization and a restricted space consumption that does not exceed the available memory. In extensive experiments on real-world diagnosis cases we compared our approaches to Reiter's HS-Tree, a state-of-the-art method that gives the same theoretical guarantees and is as general(ly applicable) as the suggested algorithms. For the computation of minimum-cardinality fault explanations, we find that (1) RBF-HS reduces memory requirements substantially in most cases by up to several orders of magnitude, (2) in more than a third of the cases, both memory savings and runtime savings are achieved, and (3) given the runtime overhead is significant, using HBF-HS instead of RBF-HS reduces the runtime to values comparable with HS-Tree while keeping the used memory reasonably bounded. When computing most probable fault explanations, we observe that RBF-HS tends to trade memory savings more or less one-to-one for runtime overheads. Again, HBF-HS proves to be a reasonable remedy to cut down the runtime while complying with practicable memory bounds.

Online solvers for partially observable Markov decision processes have difficulty scaling to problems with large action spaces. Monte Carlo tree search with progressive widening attempts to improve scaling by sampling from the action space to construct a policy search tree. The performance of progressive widening search is dependent upon the action sampling policy, often requiring problem-specific samplers. In this work, we present a general method for efficient action sampling based on Bayesian optimization. The proposed method uses a Gaussian process to model a belief over the action-value function and selects the action that will maximize the expected improvement in the optimal action value. We implement the proposed approach in a new online tree search algorithm called Bayesian Optimized Monte Carlo Planning (BOMCP). Several experiments show that BOMCP is better able to scale to large action space POMDPs than existing state-of-the-art tree search solvers.

Online solvers for partially observable Markov decision processes have difficulty scaling to problems with large action spaces. This paper proposes a method called PA-POMCPOW to sample a subset of the action space that provides varying mixtures of exploitation and exploration for inclusion in a search tree. The proposed method first evaluates the action space according to a score function that is a linear combination of expected reward and expected information gain. The actions with the highest score are then added to the search tree during tree expansion. Experiments show that PA-POMCPOW is able to outperform existing state-of-the-art solvers on problems with large discrete action spaces.

In this article, I describe agent-centered search (also called real-time search or local search) and illustrate this planning paradigm with examples. Agent-centered search methods interleave planning and plan execution and restrict planning to the part of the domain around the current state of the agent, for example, the current location of a mobile robot or the current board position of a game. These methods can execute actions in the presence of time constraints and often have a small sum of planning and execution cost, both because they trade off planning and execution cost and because they allow agents to gather information early in nondeterministic domains, which reduces the amount of planning they have to perform for unencountered situations. These advantages become important as more intelligent systems are interfaced with the world and have to operate autonomously in complex environments. Agent-centered search methods have been applied to a variety of domains, including traditional search, strips-type planning, moving-target search, planning with totally and partially observable Markov decision process models, reinforcement learning, constraint satisfaction, and robot navigation.