Algorithmic fairness has emerged as an important consideration when using machine learning to make high-stakes societal decisions. Yet, improved fairness often comes at the expense of model accuracy. While aspects of the fairness-accuracy tradeoff have been studied, most work reports the fairness and accuracy of various models separately; this makes model comparisons nearly impossible without a model-agnostic metric that reflects the balance of the two desiderata. We seek to identify, quantify, and optimize the empirical Pareto frontier of the fairness-accuracy tradeoff. Specifically, we identify and outline the empirical Pareto frontier through Tradeoff-between-Fairness-and-Accuracy (TAF) Curves; we then develop a metric to quantify this Pareto frontier through the weighted area under the TAF Curve which we term the Fairness-Area-Under-the-Curve (FAUC). TAF Curves provide the first empirical, model-agnostic characterization of the Pareto frontier, while FAUC provides the first metric to impartially compare model families on both fairness and accuracy. Both TAF Curves and FAUC can be employed with all group fairness definitions and accuracy measures. Next, we ask: Is it possible to expand the empirical Pareto frontier and thus improve the FAUC for a given collection of fitted models? We answer affirmately by developing a novel fair model stacking framework, FairStacks, that solves a convex program to maximize the accuracy of model ensemble subject to a score-bias constraint. We show that optimizing with FairStacks always expands the empirical Pareto frontier and improves the FAUC; we additionally study other theoretical properties of our proposed approach. Finally, we empirically validate TAF, FAUC, and FairStacks through studies on several real benchmark data sets, showing that FairStacks leads to major improvements in FAUC that outperform existing algorithmic fairness approaches.

Solutions to multi-objective optimization problems can generally not be compared or ordered, due to the lack of orderability of the single objectives. Furthermore, decision-makers are often made to believe that scaled objectives can be compared. This is a fallacy, as the space of solutions is in practice inhomogeneous without linear trade-offs. We present a method that uses the probability integral transform in order to map the objectives of a problem into scores that all share the same range. In the score space, we can learn which trade-offs are actually possible and develop methods for mapping the desired trade-off back into the preference space. Our results demonstrate that Pareto efficient solutions can be ordered using a low- or no-preference aggregation of the single objectives. When using scores instead of raw objectives during optimization, the process allows for obtaining trade-offs significantly closer to the expressed preference. Using a non-linear mapping for transforming a desired solution in the score space to the required preference for optimization improves this even more drastically.

Combinatorial optimization problems are complex problems with a discrete but large set of possible solutions. Some of the most renowned examples of these problems are the traveling salesman, the bin-packing, and the job-shop scheduling problems. Researchers at the Amazon Quantum Solutions Lab, part of the AWS Intelligent and Advanced Computer Technologies Labs, have recently developed a new tool to tackle combinatorial optimization problems, based on graph neural networks (GNNs). The approach developed by Schuetz, Brubaker and Katzgraber, published in Nature Machine Intelligence, could be used to optimize a variety of real-world problems. "Our work was very much inspired by customer needs," Martin Schuetz, one of the researchers who carried out the study, told TechXplore.

Abstract: This work proposes an accelerated primal-dual dynamical system for affine constrained convex optimization and presents a class of primal-dual methods with nonergodic convergence rates. In continuous level, exponential decay of a novel Lyapunov function is established and in discrete level, implicit, semi-implicit and explicit numerical discretizations for the continuous model are considered sequentially and lead to new accelerated primal-dual methods for solving linearly constrained optimization problems. Special structures of the subproblems in those schemes are utilized to develop efficient inner solvers. In addition, nonergodic convergence rates in terms of primal-dual gap, primal objective residual and feasibility violation are proved via a tailored discrete Lyapunov function. Moreover, our method has also been applied to decentralized distributed optimization for fast and efficient solution.

Mixed-integer programming (MIP) technology offers a generic way of formulating and solving combinatorial optimization problems. While generally reliable, state-of-the-art MIP solvers base many crucial decisions on hand-crafted heuristics, largely ignoring common patterns within a given instance distribution of the problem of interest. Here, we propose MIP-GNN, a general framework for enhancing such solvers with data-driven insights. By encoding the variable-constraint interactions of a given mixed-integer linear program (MILP) as a bipartite graph, we leverage state-of-the-art graph neural network architectures to predict variable biases, i.e., component-wise averages of (near) optimal solutions, indicating how likely a variable will be set to 0 or 1 in (near) optimal solutions of binary MILPs. In turn, the predicted biases stemming from a single, once-trained model are used to guide the solver, replacing heuristic components. We integrate MIP-GNN into a state-of-the-art MIP solver, applying it to tasks such as node selection and warm-starting, showing significant improvements compared to the default setting of the solver on two classes of challenging binary MILPs.

Most successful stochastic black-box optimizers, such as CMA-ES, use rankings of the individual samples to obtain a new search distribution. Yet, the use of rankings also introduces several issues such as the underlying optimization objective is often unclear, i.e., we do not optimize the expected fitness. Further, while these algorithms typically produce a high-quality mean estimate of the search distribution, the produced samples can have poor quality as these algorithms are ignorant of the regret. Lastly, noisy fitness function evaluations may result in solutions that are highly sub-optimal on expectation. In contrast, stochastic optimizers that are motivated by policy gradients, such as the Model-based Relative Entropy Stochastic Search (MORE) algorithm, directly optimize the expected fitness function without the use of rankings. MORE can be derived by applying natural policy gradients and compatible function approximation, and is using information theoretic constraints to ensure the stability of the policy update. While MORE does not suffer from the listed limitations, it often cannot achieve state of the art performance in comparison to ranking based methods. We improve MORE by decoupling the update of the mean and covariance of the search distribution allowing for more aggressive updates on the mean while keeping the update on the covariance conservative, an improved entropy scheduling technique based on an evolution path which results in faster convergence and a simplified and more effective model learning approach in comparison to the original paper. We compare our algorithm to state of the art black-box optimization algorithms on standard optimization tasks as well as on episodic RL tasks in robotics where it is also crucial to have small regret. We obtain competitive results on benchmark functions and clearly outperform ranking-based methods in terms of regret on the RL tasks.

Nonlinear trajectory optimization algorithms have been developed to handle optimal control problems with nonlinear dynamics and nonconvex constraints in trajectory planning. The performance and computational efficiency of many trajectory optimization methods are sensitive to the initial guess, i.e., the trajectory guess needed by the recursive trajectory optimization algorithm. Motivated by this observation, we tackle the initialization problem for trajectory optimization via policy optimization. To optimize a policy, we propose a guided policy search method that has two key components: i) Trajectory update; ii) Policy update. The trajectory update involves offline solutions of a large number of trajectory optimization problems from different initial states via Sequential Convex Programming (SCP). Here we take a single SCP step to generate the trajectory iterate for each problem. In conjunction with these iterates, we also generate additional trajectories around each iterate via a feedback control law. Then all these trajectories are used by a stochastic gradient descent algorithm to update the neural network policy, i.e., the policy update step. As a result, the trained policy makes it possible to generate trajectory candidates that are close to the optimality and feasibility and that provide excellent initial guesses for the trajectory optimization methods. We validate the proposed method via a real-world 6-degree-of-freedom powered descent guidance problem for a reusable rocket.

The Frank-Wolfe algorithm has seen a resurgence in popularity due to its ability to efficiently solve constrained optimization problems in machine learning and high-dimensional statistics. As such, there is much interest in establishing when the algorithm may possess a "linear" $O(\log(1/\epsilon))$ dimension-free iteration complexity comparable to projected gradient descent. In this paper, we provide a general technique for establishing domain specific and easy-to-estimate lower bounds for Frank-Wolfe and its variants using the metric entropy of the domain. Most notably, we show that a dimension-free linear upper bound must fail not only in the worst case, but in the \emph{average case}: for a Gaussian or spherical random polytope in $\mathbb{R}^d$ with $\mathrm{poly}(d)$ vertices, Frank-Wolfe requires up to $\tilde\Omega(d)$ iterations to achieve a $O(1/d)$ error bound, with high probability. We also establish this phenomenon for the nuclear norm ball. The link with metric entropy also has interesting positive implications for conditional gradient algorithms in statistics, such as gradient boosting and matching pursuit. In particular, we show that it is possible to extract fast-decaying upper bounds on the excess risk directly from an analysis of the underlying optimization procedure.

In emerging Industrial Cyber-Physical Systems (ICPSs), the joint design of communication and control sub-systems is essential, as these sub-systems are interconnected. In this paper, we study the joint design problem of an event-triggered control and an energy-efficient resource allocation in a fifth generation (5G) wireless network. We formally state the problem as a multi-objective optimization one, aiming to minimize the number of updates on the actuators' input and the power consumption in the downlink transmission. To address the problem, we propose a model-free hierarchical reinforcement learning approach \textcolor{blue}{with uniformly ultimate boundedness stability guarantee} that learns four policies simultaneously. These policies contain an update time policy on the actuators' input, a control policy, and energy-efficient sub-carrier and power allocation policies. Our simulation results show that the proposed approach can properly control a simulated ICPS and significantly decrease the number of updates on the actuators' input as well as the downlink power consumption.

Federated learning systems facilitate training of global models in settings where potentially heterogeneous data is distributed across a large number of clients. Such systems operate in settings with intermittent client availability and/or time-varying communication constraints. As a result, the global models trained by federated learning systems may be biased towards clients with higher availability. We propose F3AST, an unbiased algorithm that dynamically learns an availability-dependent client selection strategy which asymptotically minimizes the impact of client-sampling variance on the global model convergence, enhancing performance of federated learning. The proposed algorithm is tested in a variety of settings for intermittently available clients under communication constraints, and its efficacy demonstrated on synthetic data and realistically federated benchmarking experiments using CIFAR100 and Shakespeare datasets. We show up to 186% and 8% accuracy improvements over FedAvg, and 8% and 7% over FedAdam on CIFAR100 and Shakespeare, respectively.