The optimal design of many engineering processes can be subject to expensive and time-consuming experimentation. For efficiency, we seek to avoid wasting valuable resources in testing sub-optimal designs. One way to achieve this is by obtaining cheaper approximations of the desired system, which allow us to quickly explore new regimes and avoid areas that are clearly sub-optimal. As an example, consider the case diagrammed in Figure 1 from battery materials research with the goal of designing electrode materials for optimal performance in pouch cells. We can use experiments with cheaper coin cells and shorter test procedures to approximate the behaviour of the material in longer stability tests in pouch cells, which is in turn closer to the expected performance in electric car applications [Chen et al., 2019, Dörfler et al., 2020, Liu et al., 2021]. Similarly, design goals regarding battery life such as discharge capacity retention can be approximated using an early prediction model on the first few charge cycles rather than running aging and stability tests to completion [Attia et al., 2020].

Enhancing the diversity of policies is beneficial for robustness, exploration, and transfer in reinforcement learning (RL). In this paper, we aim to seek diverse policies in an under-explored setting, namely RL tasks with structured action spaces with the two properties of composability and local dependencies. The complex action structure, non-uniform reward landscape, and subtle hyperparameter tuning due to the properties of structured actions prevent existing approaches from scaling well. We propose a simple and effective RL method, Diverse Policy Optimization (DPO), to model the policies in structured action space as the energy-based models (EBM) by following the probabilistic RL framework. A recently proposed novel and powerful generative model, GFlowNet, is introduced as the efficient, diverse EBM-based policy sampler. DPO follows a joint optimization framework: the outer layer uses the diverse policies sampled by the GFlowNet to update the EBM-based policies, which supports the GFlowNet training in the inner layer. Experiments on ATSC and Battle benchmarks demonstrate that DPO can efficiently discover surprisingly diverse policies in challenging scenarios and substantially outperform existing state-of-the-art methods.

Cutting plane selection is a subroutine used in all modern mixed-integer linear programming solvers with the goal of selecting a subset of generated cuts that induce optimal solver performance. These solvers have millions of parameter combinations, and so are excellent candidates for parameter tuning. Cut selection scoring rules are usually weighted sums of different measurements, where the weights are parameters. We present a parametric family of mixed-integer linear programs together with infinitely many family-wide valid cuts. Some of these cuts can induce integer optimal solutions directly after being applied, while others fail to do so even if an infinite amount are applied. We show for a specific cut selection rule, that any finite grid search of the parameter space will always miss all parameter values, which select integer optimal inducing cuts in an infinite amount of our problems. We propose a variation on the design of existing graph convolutional neural networks, adapting them to learn cut selection rule parameters. We present a reinforcement learning framework for selecting cuts, and train our design using said framework over MIPLIB 2017 and a neural network verification data set. Our framework and design show that adaptive cut selection does substantially improve performance over a diverse set of instances, but that finding a single function describing such a rule is difficult. Code for reproducing all experiments is available at https://github.com/Opt-Mucca/Adaptive-Cutsel-MILP.

Online optimization is a well-established optimization paradigm that aims to make a sequence of correct decisions given knowledge of the correct answer to previous decision tasks. Bilevel programming involves a hierarchical optimization problem where the feasible region of the so-called outer problem is restricted by the graph of the solution set mapping of the inner problem. This paper brings these two ideas together and studies an online bilevel optimization setting in which a sequence of time-varying bilevel problems are revealed one after the other. We extend the known regret bounds for single-level online algorithms to the bilevel setting. Specifically, we introduce new notions of bilevel regret, develop an online alternating time-averaged gradient method that is capable of leveraging smoothness, and provide regret bounds in terms of the path-length of the inner and outer minimizer sequences.

Bilevel optimization plays an essential role in many machine learning tasks, ranging from hyperparameter optimization to meta-learning. Existing studies on bilevel optimization, however, focus on either centralized or synchronous distributed setting. The centralized bilevel optimization approaches require collecting a massive amount of data to a single server, which inevitably incur significant communication expenses and may give rise to data privacy risks. Synchronous distributed bilevel optimization algorithms, on the other hand, often face the straggler problem and will immediately stop working if a few workers fail to respond. As a remedy, we propose A synchronous D istributed Bilevel O ptimization (ADBO) algorithm. The proposed ADBO can tackle bilevel optimization problems with both nonconvex upper-level and lower-level objective functions, and its convergence is theoretically guaranteed. In bilevel optimization, one optimization problem is embedded or nested with another. Specifically, the outer optimization problem is called the upper-level optimization problem and the inner optimization problem is called the lower-level optimization problem.

We present a sequential hierarchical least-squares programming solver with trust-region and hierarchical step-filter tailored to prioritized non-linear optimal control. It is based on a hierarchical step-filter which resolves each priority level of a non-linear hierarchical least-squares programming via a globally convergent sequential quadratic programming step-filter. Leveraging a condition on the trust-region or the filter initialization, our hierarchical step-filter maintains this global convergence property. The hierarchical least-squares programming sub-problems are solved via a sparse nullspace method based interior point method. It is based on an efficient implementation of the turnback algorithm for the computation of nullspace bases for banded matrices. It is also here that we propose a nullspace trust region adaptation method towards a comprehensive hierarchical step-filter. We demonstrate the computational efficiency of the hierarchical solver on typical test functions like the Rosenbrock and Himmelblau's functions, inverse kinematics problems and optimal control.

In Bayesian optimisation, we often seek to minimise the black-box objective functions that arise in real-world physical systems. A primary contributor to the cost of evaluating such black-box objective functions is often the effort required to prepare the system for measurement. We consider a common scenario where preparation costs grow as the distance between successive evaluations increases. In this setting, smooth optimisation trajectories are preferred and the jumpy paths produced by the standard myopic (i.e.\ one-step-optimal) Bayesian optimisation methods are sub-optimal. Our algorithm, MONGOOSE, uses a meta-learnt parametric policy to generate smooth optimisation trajectories, achieving performance gains over existing methods when optimising functions with large movement costs.

Model Predictive Path Integral (MPPI) control is a type of sampling-based model predictive control that simulates thousands of trajectories and uses these trajectories to synthesize optimal controls on-the-fly. In practice, however, MPPI encounters problems limiting its application. For instance, it has been observed that MPPI tends to make poor decisions if unmodeled dynamics or environmental disturbances exist, preventing its use in safety-critical applications. Moreover, the multi-threaded simulations used by MPPI require significant onboard computational resources, making the algorithm inaccessible to robots without modern GPUs. To alleviate these issues, we propose a novel (Shield-MPPI) algorithm that provides robustness against unpredicted disturbances and achieves real-time planning using a much smaller number of parallel simulations on regular CPUs. The novel Shield-MPPI algorithm is tested on an aggressive autonomous racing platform both in simulation and using experiments. The results show that the proposed controller greatly reduces the number of constraint violations compared to state-of-the-art robust MPPI variants and stochastic MPC methods.

A recourse action aims to explain a particular algorithmic decision by showing one specific way in which the instance could be modified to receive an alternate outcome. Existing recourse generation methods often assume that the machine learning model does not change over time. However, this assumption does not always hold in practice because of data distribution shifts, and in this case, the recourse action may become invalid. To redress this shortcoming, we propose the Distributionally Robust Recourse Action (DiRRAc) framework, which generates a recourse action that has a high probability of being valid under a mixture of model shifts. We formulate the robustified recourse setup as a min-max optimization problem, where the max problem is specified by Gelbrich distance over an ambiguity set around the distribution of model parameters. Then we suggest a projected gradient descent algorithm to find a robust recourse according to the min-max objective. We show that our DiRRAc framework can be extended to hedge against the misspecification of the mixture weights. Numerical experiments with both synthetic and three real-world datasets demonstrate the benefits of our proposed framework over state-of-the-art recourse methods.

One of the most crucial and challenging steps in the development of robots intended to restore the mobility of the human body after a loss of functional movement due to neurological injuries is the IK of physiological limbs, which consists of computing joint angles configuration based on the predefined input workspace coordinates. Generally speaking, the complexity of the IK problem depends on the geometry of the manipulator and the nonlinearity of its model, which gives the corresponding relation between the task and the joint spaces. Furthermore, IK solution is essential for the real-time control. Thus, it must be precise in order to enable the robot to perform the task successfully. IK techniques can be classified into three categories, namely, analytical method, numerical method, and intelligent method. The analytical method solves IK by solving a set of closed-form equations that can give the generalized coordinate value that drives the end effector of the manipulator to the predefined target position [1].