In deep learning, stochastic gradient descent (SGD) and its momentum-based variants are widely used in optimization algorithms, they usually face the problem of slow convergence. Meanwhile, existing adaptive learning rate optimizers accelerate convergence but often at the expense of generalization ability. We demonstrate that the adaptive learning rate property impairs generalization. To address this contradiction, we propose a novel optimization method that aims to accelerate the convergence rate of SGD without loss of generalization. This approach is based on the idea of reducing the variance of the historical gradient, enhancing the first-order moment estimation of the SGD by applying Wiener filtering theory, and introducing a time-varying adaptive weight. Experimental results show that SGDF achieves a trade-off between convergence and generalization compared to state-of-the-art optimizers.

We present a randomized forward mode gradient (RFG) as an alternative to backpropagation. RFG is a random estimator for the gradient that is constructed based on the directional derivative along a random vector. The forward mode automatic differentiation (AD) provides an efficient computation of RFG. The probability distribution of the random vector determines the statistical properties of RFG. Through the second moment analysis, we found that the distribution with the smallest kurtosis yields the smallest expected relative squared error. By replacing gradient with RFG, a class of RFG-based optimization algorithms is obtained. By focusing on gradient descent (GD) and Polyak's heavy ball (PHB) methods, we present a convergence analysis of RFG-based optimization algorithms for quadratic functions. Computational experiments are presented to demonstrate the performance of the proposed algorithms and verify the theoretical findings.

Decision Transformer (DT), which employs expressive sequence modeling techniques to perform action generation, has emerged as a promising approach to offline policy optimization. However, DT generates actions conditioned on a desired future return, which is known to bear some weaknesses such as the susceptibility to environmental stochasticity. To overcome DT's weaknesses, we propose to empower DT with dynamic programming. Our method comprises three steps. First, we employ in-sample value iteration to obtain approximated value functions, which involves dynamic programming over the MDP structure. Second, we evaluate action quality in context with estimated advantages. We introduce two types of advantage estimators, IAE and GAE, which are suitable for different tasks. Third, we train an Advantage-Conditioned Transformer (ACT) to generate actions conditioned on the estimated advantages. Finally, during testing, ACT generates actions conditioned on a desired advantage. Our evaluation results validate that, by leveraging the power of dynamic programming, ACT demonstrates effective trajectory stitching and robust action generation in spite of the environmental stochasticity, outperforming baseline methods across various benchmarks. Additionally, we conduct an in-depth analysis of ACT's various design choices through ablation studies. Our code is available at https://github.com/LAMDA-RL/ACT.

The 5G mobile networks extend the capability for supporting collaborative robot operations in outdoor scenarios. However, the restricted battery life of robots still poses a major obstacle to their effective implementation and utilization in real scenarios. One of the most challenging situations is the execution of mission-critical tasks that require the use of various onboard sensors to perform simultaneous localization and mapping (SLAM) of unexplored environments. Given the time-sensitive nature of these tasks, completing them in the shortest possible time is of the highest importance. In this paper, we analyze the benefits of 5G-enabled collaborative robots by enhancing the intelligence of the robot operation through joint orchestration of Robot Operating System (ROS) and 5G resources for energysaving goals, addressing the problem from both offline and online manners. We propose OROS, a novel orchestration approach that minimizes mission-critical task completion times as well as overall energy consumption of 5G-connected robots by jointly optimizing robotic navigation and sensing together with infrastructure resources. We validate our 5G-enabled collaborative framework by means of Matlab/Simulink, ROS software and Gazebo simulator. Our results show an improvement between 3.65% and 11.98% in exploration task by exploiting 5G orchestration features for battery savings when using 3 robots.

Policy-gradient algorithms are effective reinforcement learning methods for solving control problems with continuous state and action spaces. To compute near-optimal policies, it is essential in practice to include exploration terms in the learning objective. Although the effectiveness of these terms is usually justified by an intrinsic need to explore environments, we propose a novel analysis and distinguish two different implications of these techniques. First, they make it possible to smooth the learning objective and to eliminate local optima while preserving the global maximum. Second, they modify the gradient estimates, increasing the probability that the stochastic parameter update eventually provides an optimal policy. In light of these effects, we discuss and illustrate empirically exploration strategies based on entropy bonuses, highlighting their limitations and opening avenues for future works in the design and analysis of such strategies.

The convergence of expectation-maximization (EM)-based algorithms typically requires continuity of the likelihood function with respect to all the unknown parameters (optimization variables). The requirement is not met when parameters comprise both discrete and continuous variables, making the convergence analysis nontrivial. This paper introduces a set of conditions that ensure the convergence of a specific class of EM algorithms that estimate a mixture of discrete and continuous parameters. Our results offer a new analysis technique for iterative algorithms that solve mixed-integer non-linear optimization problems. As a concrete example, we prove the convergence of the EM-based sparse Bayesian learning algorithm in [1] that estimates the state of a linear dynamical system with jointly sparse inputs and bursty missing observations. Our results establish that the algorithm in [1] converges to the set of stationary points of the maximum likelihood cost with respect to the continuous optimization variables.

We introduce AlphaRank, an artificial intelligence approach to address the fixed-budget ranking and selection (R&S) problems. We formulate the sequential sampling decision as a Markov decision process and propose a Monte Carlo simulation-based rollout policy that utilizes classic R&S procedures as base policies for efficiently learning the value function of stochastic dynamic programming. We accelerate online sample-allocation by using deep reinforcement learning to pre-train a neural network model offline based on a given prior. We also propose a parallelizable computing framework for large-scale problems, effectively combining "divide and conquer" and "recursion" for enhanced scalability and efficiency. Numerical experiments demonstrate that the performance of AlphaRank is significantly improved over the base policies, which could be attributed to AlphaRank's superior capability on the trade-off among mean, variance, and induced correlation overlooked by many existing policies.

Submodular functions, as well as the sub-class of decomposable submodular functions, and their optimization appear in a wide range of applications in machine learning, recommendation systems, and welfare maximization. However, optimization of decomposable submodular functions with millions of component functions is computationally prohibitive. Furthermore, the component functions may be private (they might represent user preference function, for example) and cannot be widely shared. To address these issues, we propose a {\em federated optimization} setting for decomposable submodular optimization. In this setting, clients have their own preference functions, and a weighted sum of these preferences needs to be maximized. We implement the popular {\em continuous greedy} algorithm in this setting where clients take parallel small local steps towards the local solution and then the local changes are aggregated at a central server. To address the large number of clients, the aggregation is performed only on a subsampled set. Further, the aggregation is performed only intermittently between stretches of parallel local steps, which reduces communication cost significantly. We show that our federated algorithm is guaranteed to provide a good approximate solution, even in the presence of above cost-cutting measures. Finally, we show how the federated setting can be incorporated in solving fundamental discrete submodular optimization problems such as Maximum Coverage and Facility Location.

Motivated by COVID-19 vaccine allocation, where vulnerable subpopulations are simultaneously more impacted in terms of health and more disadvantaged in terms of access to the vaccine, we formalize and study the problem of resource allocation when there are inherent access differences that correlate with advantage and disadvantage. We identify reducing resource disparity as a key goal in this context and show its role as a proxy to more nuanced downstream impacts. We develop a concrete access model that helps quantify how a given allocation translates to resource flow for the advantaged vs. the disadvantaged, based on the access gap between them. We then provide a methodology for access-aware allocation. Intuitively, the resulting allocation leverages more vaccines in locations with higher vulnerable populations to mitigate the access gap and reduce overall disparity. Surprisingly, knowledge of the access gap is often not needed to perform access-aware allocation. To support this formalism, we provide empirical evidence for our access model and show that access-aware allocation can significantly reduce resource disparity and thus improve downstream outcomes. We demonstrate this at various scales, including at county, state, national, and global levels.

Machine teaching often involves the creation of an optimal (typically minimal) dataset to help a model (referred to as the `student') achieve specific goals given by a teacher. While abundant in the continuous domain, the studies on the effectiveness of machine teaching in the discrete domain are relatively limited. This paper focuses on machine teaching in the discrete domain, specifically on manipulating student models' predictions based on the goals of teachers via changing the training data efficiently. We formulate this task as a combinatorial optimization problem and solve it by proposing an iterative searching algorithm. Our algorithm demonstrates significant numerical merit in the scenarios where a teacher attempts at correcting erroneous predictions to improve the student's models, or maliciously manipulating the model to misclassify some specific samples to the target class aligned with his personal profits. Experimental results show that our proposed algorithm can have superior performance in effectively and efficiently manipulating the predictions of the model, surpassing conventional baselines.