Bayesian optimization has emerged at the forefront of expensive black-box optimization due to its data efficiency. Recent years have witnessed a proliferation of studies on the development of new Bayesian optimization algorithms and their applications. Hence, this paper attempts to provide a comprehensive and updated survey of recent advances in Bayesian optimization and identify interesting open problems. We categorize the existing work on Bayesian optimization into nine main groups according to the motivations and focus of the proposed algorithms. For each category, we present the main advances with respect to the construction of surrogate models and adaptation of the acquisition functions. Finally, we discuss the open questions and suggest promising future research directions, in particular with regard to heterogeneity, privacy preservation, and fairness in distributed and federated optimization systems.

The DEtection TRansformer (DETR) approach, which uses a transformer encoder-decoder architecture and a set-based global loss, has become a building block in many transformer based applications. However, as originally presented, the assignment cost and the global loss are not aligned, i.e., reducing the former is likely but not guaranteed to reduce the latter. And the issue of gradient is ignored when a combinatorial solver such as Hungarian is used. In this paper we show that the global loss can be expressed as the sum of an assignment-independent term, and an assignment-dependent term which can be used to define the assignment cost matrix. Recent results on generalized gradients of optimal assignment cost with respect to parameters of an assignment problem are then used to define generalized gradients of the loss with respect to network parameters, and backpropagation is carried out properly. Our experiments using the same loss weights show interesting convergence properties and a potential for further performance improvements.

The Facility for Rare Isotope Beams (FRIB) at Michigan State University (MSU) is designed for various kinds of rare isotope production. This involves the frequent switch of the ion source species. Therefore, fast tuning of the accelerator Front-End (FE) to maintain optimal beam optics is one of the key performance requirements. Breaking-through the tuning performance over the traditional black-box optimization algorithm may be possible if historical or simulated data can be incorporated into the optimization algorithm in a computationally feasible way. However, we experienced significant machine status changes (a.k.a. 'distribution shift' or'machine drift') whenever ion source species are switched or the ion source is re-started (after overnight turn-off).

For certain diseases or traits, genes play a crucial role in their expression. And the interactions between genes, i.e., gene regulatory networks (GRNs), have become a recent research hotspot. With the availability of high-throughput gene expression data and the substantial increase in arithmetic power, it is possible to reconstruct large-scale gene regulatory networks[1]. Due to the nature of gene expression data, providing information about the abundance of mRNAs only rather than binding information, gene regulatory networks defined in the above sense provide information about regulatory interactions between regulators and their potential targets; genegene interactions, and potential protein-protein interactions. In this paper, we refer to the network inferred in this way as a gene regulatory network[32]. In gene regulatory networks, genes can be divided into two categories. Transcription factors (TF), also known as trans-acting factors, are DNA-binding proteins that specifically interact with the cis-acting elements of eukaryotic genes and have an activating or inhibiting effect on gene transcription. The gene that receives this activation or repression is referred to as the target gene. It is important to note that transcription factors themselves may also be target genes, i.e., there may be mutual regulation in the regulatory network.

Soft robots are interesting examples of hyper-redundancy in robotics, however, the nonlinear continuous dynamics of these robots and the use of hyper-elastic and visco-elastic materials makes modeling of these robots more complicated. This study presents a geometric Inverse Kinematic (IK) model for trajectory tracking of multi-segment extensible soft robots, where, each segment of the soft actuator is geometrically approximated with multiple rigid links connected with rotary and prismatic joints. Using optimization methods, the desired configuration variables of the soft actuator for the desired end-effector positions are obtained. Also, the redundancy of the robot is applied for second task applications, such as tip angle control. The model's performance is investigated through simulations, numerical benchmarks, and experimental validations and results show lower computational costs and higher accuracy compared to most existing methods. The method is easy to apply to multi segment soft robots, both in 2D and 3D. As a case study, a fully 3D-printed soft robot manipulator is tested using a control unit and the model predictions show good agreement with the experimental results.

This paper focuses on stochastic saddle point problems with decision-dependent distributions. These are problems whose objective is the expected value of a stochastic payoff function and whose data distribution drifts in response to decision variables--a phenomenon represented by a distributional map. A common approach to accommodating distributional shift is to retrain optimal decisions once a new distribution is revealed, or repeated retraining. We introduce the notion of equilibrium points, which are the fixed points of this repeated retraining procedure, and provide sufficient conditions for their existence and uniqueness. To find equilibrium points, we develop deterministic and stochastic primal-dual algorithms and demonstrate their convergence with constant step-size in the former and polynomial decay step-size schedule in the latter. By modeling errors emerging from a stochastic gradient estimator as sub-Weibull random variables, we provide error bounds in expectation and in high probability that hold for each iteration. Without additional knowledge of the distributional map, computing saddle points is intractable. Thus we propose a condition on the distributional map--which we call opposing mixture dominance--that ensures that the objective is strongly-convex-strongly-concave. Finally, we demonstrate that derivative-free algorithms with a single function evaluation are capable of approximating saddle points

Backpropagation applied to handwritten zip code recognition.

Countless signal processing applications include the reconstruction of signals from few indirect linear measurements. The design of effective measurement operators is typically constrained by the underlying hardware and physics, posing a challenging and often even discrete optimization task. While the potential of gradient-based learning via the unrolling of iterative recovery algorithms has been demonstrated, it has remained unclear how to leverage this technique when the set of admissible measurement operators is structured and discrete. We tackle this problem by combining unrolled optimization with Gumbel reparametrizations, which enable the computation of low-variance gradient estimates of categorical random variables. Our approach is formalized by GLODISMO (Gradient-based Learning of DIscrete Structured Measurement Operators). This novel method is easy-to-implement, computationally efficient, and extendable due to its compatibility with automatic differentiation. We empirically demonstrate the performance and flexibility of GLODISMO in several prototypical signal recovery applications, verifying that the learned measurement matrices outperform conventional designs based on randomization as well as discrete optimization baselines.

Offline goal-conditioned reinforcement learning (GCRL) promises general-purpose skill learning in the form of reaching diverse goals from purely offline datasets. We propose $\textbf{Go}$al-conditioned $f$-$\textbf{A}$dvantage $\textbf{R}$egression (GoFAR), a novel regression-based offline GCRL algorithm derived from a state-occupancy matching perspective; the key intuition is that the goal-reaching task can be formulated as a state-occupancy matching problem between a dynamics-abiding imitator agent and an expert agent that directly teleports to the goal. In contrast to prior approaches, GoFAR does not require any hindsight relabeling and enjoys uninterleaved optimization for its value and policy networks. These distinct features confer GoFAR with much better offline performance and stability as well as statistical performance guarantee that is unattainable for prior methods. Furthermore, we demonstrate that GoFAR's training objectives can be re-purposed to learn an agent-independent goal-conditioned planner from purely offline source-domain data, which enables zero-shot transfer to new target domains. Through extensive experiments, we validate GoFAR's effectiveness in various problem settings and tasks, significantly outperforming prior state-of-art. Notably, on a real robotic dexterous manipulation task, while no other method makes meaningful progress, GoFAR acquires complex manipulation behavior that successfully accomplishes diverse goals.

In this letter, we study the problem of accelerating reaching average consensus over connected graphs in a discrete-time communication setting. Literature has shown that consensus algorithms can be accelerated by increasing the graph connectivity or optimizing the weights agents place on the information received from their neighbors. Here, instead of altering the communication graph, we investigate two methods that use buffered states to accelerate reaching average consensus over a given graph. In the first method, we study how convergence rate of the well-known first-order Laplacian average consensus algorithm changes when agreement feedback is generated from buffered states. For this study, we obtain a sufficient condition on the ranges of buffered state that leads to faster convergence. In the second proposed method, we show how the average consensus problem can be cast as a convex optimization problem and solved by first-order accelerated optimization algorithms for strongly-convex cost functions. We construct an accelerated average consensus algorithm using the so-called Triple Momentum optimization algorithm. The first approach requires less global knowledge for choosing the step size, whereas the second one converges faster in our numerical results by using extra information from the graph topology. We demonstrate our results by implementing the proposed algorithms in a Gaussian Mixture Model (GMM) estimation problem used in sensor networks.