We study a class of multi-period online decision-making problems with sequence-based predictions, which may be generated by machine learning models but whose accuracy is not guaranteed. In each period, the decision-maker observes the realized request and must take an irrevocable action that yields a reward or incurs a cost, without knowledge of future arrivals. We introduce a bounded-influence framework, in which past decisions and requests exert only limited impact on the future optimal reward. Within this framework, we propose the AdaSwitch meta-algorithm, which exploits predictions to attain performance close to the offline benchmark when predictions are accurate, while preserving classical competitive-ratio guarantees under highly inaccurate predictions. Our framework and meta-algorithm apply to diverse settings, including lead-time quotation in processing systems, the $k$-server problem, and online allocation of reusable resources. These applications illustrate the flexibility and broad applicability of our approach to learning-augmented online decision-making.

The performance of industrial robotic work cells depends on optimizing various hyperparameters referring to the cell layout, such as robot base placement, tool placement, and kinematic design. Achieving this requires a bilevel optimization approach, where the high-level optimization adjusts these hyperparameters, and the low-level optimization computes robot motions. However, computing the optimal robot motion is computationally infeasible, introducing trade-offs in motion planning to make the problem tractable. These trade-offs significantly impact the overall performance of the bilevel optimization, but their effects still need to be systematically evaluated. In this paper, we introduce metrics to assess these trade-offs regarding optimality, time gain, robustness, and consistency. Through extensive simulation studies, we investigate how simplifications in motion-level optimization affect the high-level optimization outcomes, balancing computational complexity with solution quality. The proposed algorithms are applied to find the time-optimal kinematic design for a modular robot in two palletization scenarios.

Machine learning models require datasets for effective training, but directly sharing raw data poses significant privacy risk such as membership inference attacks (MIA). To mitigate the risk, privacy-preserving techniques such as data perturbation, generalization, and synthetic data generation are commonly utilized. However, these methods often degrade data accuracy, specificity, and diversity, limiting the performance of downstream tasks and thus reducing data utility. Therefore, striking an optimal balance between privacy preservation and data utility remains a critical challenge. To address this issue, we introduce a novel bilevel optimization framework for the publication of private datasets, where the upper-level task focuses on data utility and the lower-level task focuses on data privacy. In the upper-level task, a discriminator guides the generation process to ensure that perturbed latent variables are mapped to high-quality samples, maintaining fidelity for downstream tasks. In the lower-level task, our framework employs local extrinsic curvature on the data manifold as a quantitative measure of individual vulnerability to MIA, providing a geometric foundation for targeted privacy protection. By perturbing samples toward low-curvature regions, our method effectively suppresses distinctive feature combinations that are vulnerable to MIA. Through alternating optimization of both objectives, we achieve a synergistic balance between privacy and utility. Extensive experimental evaluations demonstrate that our method not only enhances resistance to MIA in downstream tasks but also surpasses existing methods in terms of sample quality and diversity.

--Appointment scheduling is a great challenge in healthcare operations management. Appointment rules (AR) provide medical practitioners with a simple yet effective tool to determine patient appointment times. Genetic programming (GP) can be used to evolve ARs. However, directly applying GP to design ARs may lead to rules that are difficult for end-users to interpret and trust. A key reason is that GP is unaware of the dimensional consistency, which ensures that the evolved rules align with users' domain knowledge and intuitive understanding. In this paper, we develop a new dimensionally aware GP algorithm with dimension repair to evolve ARs with dimensional consistency and high performance. A key innovation of our method is the dimension repair procedure, which optimizes the dimensional consistency of an expression tree while minimizing structural changes and ensuring that its output dimension meets the problem's requirements. We formulate the task as a mixed-integer linear programming model that can be efficiently solved using common mathematical programming methods. With the support of the dimension repair procedure, our method can explore a wider range of AR structures by temporarily breaking the dimensional consistency of individuals, and then restoring it without altering their overall structure, thereby identifying individuals with greater potential advantages. We evaluated the proposed method in a comprehensive set of simulated clinics. The experimental results demonstrate that our approach managed to evolve high-quality ARs that significantly outperform not only the manually designed ARs but also existing state-of-the-art dimensionally aware GP methods in terms of both objective values and dimensional consistency. In addition, we analyzed the semantics of the evolved ARs, providing insight into the design of more effective and interpretable ARs. PPOINTMENT scheduling plays a crucial role in healthcare systems, impacting clinical, operational, and financial performance. A well-designed appointment system can improve the efficiency of medical providers and patient satisfaction by smoothing demand and mitigating uncertainty in patient arrivals [1]. Patients expect fast and timely service and find long wait times increasingly difficult to tolerate. Medical providers face pressure to efficiently utilize resources, including doctors' availability and expensive diagnostic machines.

Quantum annealers provide an effective framework for solving large-scale combinatorial optimization problems. This work presents a novel methodology for training Variational Quantum Algorithms (VQAs) by reformulating the parameter optimization task as a Quadratic Unconstrained Binary Optimization (QUBO) problem. Unlike traditional gradient-based methods, our approach directly leverages the Hamiltonian of the chosen VQA ansatz and employs an adaptive, metaheuristic optimization scheme. This optimization strategy provides a rich set of configurable parameters which enables the adaptation to specific problem characteristics and available computational resources. The proposed framework is generalizable to arbitrary Hamiltonians and integrates a recursive refinement strategy to progressively approximate high-quality solutions. Experimental evaluations demonstrate the feasibility of the method and its ability to significantly reduce computational overhead compared to classical and evolutionary optimizers, while achieving comparable or superior solution quality. These findings suggest that quantum annealers can serve as a scalable alternative to classical optimizers for VQA training, particularly in scenarios affected by barren plateaus and noisy gradient estimates, and open new possibilities for hybrid quantum gate - quantum annealing - classical optimization models in near-term quantum computing.

Efficient robot control often requires balancing task performance with energy expenditure. A common approach in reinforcement learning (RL) is to penalize energy use directly as part of the reward function. This requires carefully tuning weight terms to avoid undesirable trade-offs where energy minimization harms task success. In this work, we propose a hyperparameter-free gradient optimization method to minimize energy expenditure without conflicting with task performance. Inspired by recent works in multitask learning, our method applies policy gradient projection between task and energy objectives to derive policy updates that minimize energy expenditure in ways that do not impact task performance. We evaluate this technique on standard locomotion benchmarks of DM-Control and HumanoidBench and demonstrate a reduction of 64% energy usage while maintaining comparable task performance. Further, we conduct experiments on a Unitree GO2 quadruped showcasing Sim2Real transfer of energy efficient policies. Our method is easy to implement in standard RL pipelines with minimal code changes, is applicable to any policy gradient method, and offers a principled alternative to reward shaping for energy efficient control policies.

Visual SLAM has regained attention due to its ability to provide perceptual capabilities and simulation test data for Embodied AI. However, traditional SLAM methods struggle to meet the demands of high-quality scene reconstruction, and Gaussian SLAM systems, despite their rapid rendering and high-quality mapping capabilities, lack effective pose optimization methods and face challenges in geometric reconstruction. To address these issues, we introduce FGO-SLAM, a Gaussian SLAM system that employs an opacity radiance field as the scene representation to enhance geometric mapping performance. After initial pose estimation, we apply global adjustment to optimize camera poses and sparse point cloud, ensuring robust tracking of our approach. Additionally, we maintain a globally consistent opacity radiance field based on 3D Gaussians and introduce depth distortion and normal consistency terms to refine the scene representation. Furthermore, after constructing tetrahedral grids, we identify level sets to directly extract surfaces from 3D Gaussians. Results across various real-world and large-scale synthetic datasets demonstrate that our method achieves state-of-the-art tracking accuracy and mapping performance.

The recent development of Large Language Models (LLMs) has been accompanied by an effervescence of novel ideas and methods to better optimize the loss of deep learning models. Claims from those methods are myriad: from faster convergence to removing reliance on certain hyperparameters. However, the diverse experimental protocols used to validate these claims make direct comparisons between methods challenging. This study presents a comprehensive evaluation of recent optimization techniques across standardized LLM pretraining scenarios, systematically varying model size, batch size, and training duration. Through careful tuning of each method, we provide guidance to practitioners on which optimizer is best suited for each scenario. For researchers, our work highlights promising directions for future optimization research. Finally, by releasing our code and making all experiments fully reproducible, we hope our efforts can help the development and rigorous benchmarking of future methods.

--Community detection in graphs is crucial for understanding the organization of nodes into densely connected clusters. While numerous strategies have been developed to identify these clusters, the success of community detection can lead to privacy and information security concerns, as individuals may not want their personal information exposed. T o address this, community deception methods have been proposed to reduce the effectiveness of detection algorithms. Nevertheless, several limitations, such as the rationality of evaluation metrics and the unnoticeability of attacks, have been ignored in current deception methods. Therefore, in this work, we first investigate the limitations of the widely used deception metric, i.e., the decrease of modularity, through empirical studies. Then, we propose a new deception metric, and combine this new metric together with the attack budget to model the unnoticeable community deception task as a multi-objective optimization problem. T o further improve the deception performance, we propose two variant methods by incorporating the degree-biased and community-biased candidate node selection mechanisms. Extensive experiments on three benchmark datasets demonstrate the superiority of the proposed community deception strategies. RAPHS or networks, which consist of a series of nodes and links, have been widely employed to model complex relationships in modern social systems. The typical examples of networks include social networks which illustrate the social relationships between people, financial networks which represent the transaction records between accounts, power networks which indicate the transmission relationship between power nodes, among others [1]-[4]. In these years, significant efforts have been put into investigating the abundant information behind graph-based systems, such as community detection for clustering nodes in the graph into different groups, node classification for identifying the labels of nodes, link prediction for predicting possible future connections between nodes, etc [5]-[7]. Specifically, in this work, we focus on the task of community detection. Junyuan Fang and Chi K. Tse are with the Department of Electrical Engineering, City University of Hong Kong, Hong Kong SAR, China. Huimin Liu and Y ueqi Peng are with the School of Computer Science and Engineering, Sun Y at-sen University, Guangzhou 510006, China.

Reward maximization, safe exploration, and intrinsic motivation are often studied as separate objectives in reinforcement learning (RL). We present a unified geometric framework, that views these goals as instances of a single optimization problem on the space of achievable long-term behavior in an environment. Within this framework, classical methods such as policy mirror descent, natural policy gradient, and trust-region algorithms naturally generalize to nonlinear utilities and convex constraints. We illustrate how this perspective captures robustness, safety, exploration, and diversity objectives, and outline open challenges at the interface of geometry and deep RL.