Bayesian optimization provides sample-efficient global optimization for a broad range of applications, including automatic machine learning, molecular chemistry, and experimental design. We introduce BoTorch, a modern programming framework for Bayesian optimization. Enabled by Monte-Carlo (MC) acquisition functions and auto-differentiation, BoTorch's modular design facilitates flexible specification and optimization of probabilistic models written in PyTorch, radically simplifying implementation of novel acquisition functions. Our MC approach is made practical by a distinctive algorithmic foundation that leverages fast predictive distributions and hardware acceleration. In experiments, we demonstrate the improved sample efficiency of BoTorch relative to other popular libraries. BoTorch is open source and available at https://github.com/pytorch/botorch.

Federated learning is a key scenario in modern large-scale m achine learning. In that scenario, the training data remains distributed over a larg e number of clients, which may be phones, other mobile devices, or network sensors and a centr alized model is learned without ever transmitting client data over the network. The standar d optimization algorithm used in this scenario is Federated A veraging (FedA vg). However, when client data is heterogeneous, which is typical in applications, FedA vg does not a dmit a favorable convergence guarantee. This is because local updates on clients can drif t apart, which also explains the slow convergence and hard-to-tune nature of FedA vg in pract ice. This paper presents a new Stochastic Controlled A veraging algorithm ( SCAFFOLD) which uses control variates to reduce the drift between different clients. We prove that the algorithm requires significantly fewer rounds of communication and benefits from favorable co nvergence guarantees.

We present multi-point optimization: an optimization technique that allows to train several models simultaneously without the need to keep the parameters of each one individually. The proposed method is used for a thorough empirical analysis of the loss landscape of neural networks. By extensive experiments on FashionMNIST and CIFAR10 datasets we demonstrate two things: 1) loss surface is surprisingly diverse and intricate in terms of landscape patterns it contains, and 2) adding batch normalization makes it more smooth. Source code to reproduce all the reported results is available on GitHub: https://github.com/universome/loss-patterns.

To integrate strategic, tactical and operational decisions, the two-stage optimization has been widely used to guide dynamic decision making. In this paper, we study the two-stage stochastic programming for complex systems with unknown response estimated by simulation. We introduce the global-local metamodel assisted two-stage optimization via simulation that can efficiently employ the simulation resource to iteratively solve for the optimal first- and second-stage decisions. Specifically, at each visited first-stage decision, we develop a local metamodel to simultaneously solve a set of scenario-based second-stage optimization problems, which also allows us to estimate the optimality gap. Then, we construct a global metamodel accounting for the errors induced by: (1) using a finite number of scenarios to approximate the expected future cost occurring in the planning horizon, (2) second-stage optimality gap, and (3) finite visited first-stage decisions. Assisted by the global-local metamodel, we propose a new simulation optimization approach that can efficiently and iteratively search for the optimal first- and second-stage decisions. Our framework can guarantee the convergence of optimal solution for the discrete two-stage optimization with unknown objective, and the empirical study indicates that it achieves substantial efficiency and accuracy.

Discrete integration in a high dimensional space of $n$ variables poses fundamental challenges. The WISH algorithm reduces the intractable discrete integration problem into $n$ optimization queries subject to randomized constraints, obtaining a constant approximation guarantee. The optimization queries are expensive, which limits the applicability of WISH. We propose AdaWISH, which is able to obtain the same guarantee, but accesses only a small subset of queries of WISH. For example, when the number of function values is bounded by a constant, AdaWISH issues only $O(\log n)$ queries. The key idea is to query adaptively, taking advantage of the shape of the weight function. In general, we prove that AdaWISH has a regret of no more than $O(\log n)$ relative to an oracle that issues queries at data-dependent optimal points. Experimentally, AdaWISH gives precise estimates for discrete integration problems, of the same quality as that of WISH and better than several competing approaches, on a variety of probabilistic inference benchmarks, while saving substantially on the number of optimization queries compared to WISH. For example, it saves $81.5\%$ of WISH queries while retaining the quality of results on a suite of UAI inference challenge benchmarks.

We present a preference learning framework for multiple criteria sorting. We consider sorting procedures applying an additive value model with diverse types of marginal value functions (including linear, piecewise-linear, splined, and general monotone ones) under a unified analytical framework. Differently from the existing sorting methods that infer a preference model from crisp decision examples, where each reference alternative is assigned to a unique class, our framework allows to consider valued assignment examples in which a reference alternative can be classified into multiple classes with respective credibility degrees. We propose an optimization model for constructing a preference model from such valued examples by maximizing the credible consistency among reference alternatives. To improve the predictive ability of the constructed model on new instances, we employ the regularization techniques. Moreover, to enhance the capability of addressing large-scale datasets, we introduce a state-of-the-art algorithm that is widely used in the machine learning community to solve the proposed optimization model in a computationally efficient way. Using the constructed additive value model, we determine both crisp and valued assignments for non-reference alternatives. Moreover, we allow the Decision Maker to prioritize importance of classes and give the method a flexibility to adjust classification performance across classes according to the specified priorities. The practical usefulness of the analytical framework is demonstrated on a real-world dataset by comparing it to several existing sorting methods.

Despite an ever growing literature on reinforcement learning algorithms and applications, much less is known about their statistical inference. In this paper, we investigate the large sample behaviors of the Q-value estimates with closed-form characterizations of the asymptotic variances. This allows us to efficiently construct confidence regions for Q-value and optimal value functions, and to develop policies to minimize their estimation errors. This also leads to a policy exploration strategy that relies on estimating the relative discrepancies among the Q estimates. Numerical experiments show superior performances of our exploration strategy than other benchmark approaches.

Optimal transport distances are powerful tools to compare probability distributions and have found many applications in machine learning. Yet their algorithmic complexity prevents their direct use on large scale datasets. To overcome this challenge, practitioners compute these distances on minibatches {\em i.e.} they average the outcome of several smaller optimal transport problems. We propose in this paper an analysis of this practice, which effects are not well understood so far. We notably argue that it is equivalent to an implicit regularization of the original problem, with appealing properties such as unbiased estimators, gradients and a concentration bound around the expectation, but also with defects such as loss of distance property. Along with this theoretical analysis, we also conduct empirical experiments on gradient flows, GANs or color transfer that highlight the practical interest of this strategy.

The job sequencing and tool switching problem (SSP) has been extensively studied in the field of operations research, due to its practical relevance and methodological interest. Given a machine that can load a limited amount of tools simultaneously and a number of jobs that require a subset of the available tools, the SSP seeks a job sequence that minimizes the number of tool switches in the machine. To solve this problem, we propose a simple and efficient hybrid genetic search based on a generic solution representation, a tailored decoding operator, efficient local searches and diversity management techniques. To guide the search, we introduce a secondary objective designed to break ties. These techniques allow to explore structurally different solutions and escape local optima. As shown in our computational experiments on classical benchmark instances, our algorithm significantly outperforms all previous approaches while remaining simple to apprehend and easy to implement. We finally report results on a new set of larger instances to stimulate future research and comparative analyses.

In management, business, economics, scien ce, engineering, and research domains, L arge Scale Global Optimization (LSGO) plays a predominant and vital role. Though LSGO is applied in many of the application domains, it is a very troublesome and a perverse task . The Congress o n Evolutionary Comp utation (CEC) began a n LSGO competition to come up with algorithms with a bunch of standard benchmark unconstrained LS GO functions . Therefore, in this paper, we propos e a hybrid meta - heuristic algorithm, which combines a n I mproved and M odified Harmony Search (IMHS), along with a M odified Differential Evolution (MDE) with an alternate selection strategy . Harmony Search (HS) does the job of exploration and exploitation, and Differe ntial Evolution does the job of giving a perturbation to the exploration of IMHS, as harmony search suffers from being stuck at the basin of local optimal . To judge the performance of the suggested algorithm, we compare the proposed algorithm with ten excellent met a - heuristic algorithms on fifteen LSGO benchmark functions, which have 1000 continuous decision variables, of the CEC 2013 LSGO special session . The experimental results consistently show that our proposed hybrid meta - heuristic performs statistically on par with some algorithms in a few problems, while it turned out to be the best in a couple of problems.