In this paper, we introduce a versatile scheme for optimizing the arrival rates of quasi-reversible queueing systems. We first propose an alternative definition of quasi-reversibility that encompasses reversibility and highlights the importance of the definition of customer classes. In a second time, we introduce balanced arrival control policies, which generalize the notion of balanced arrival rates introduced in the context of Whittle networks, to the much broader class of quasi-reversible queueing systems. We prove that supplementing a quasi-reversible queueing system with a balanced arrival-control policy preserves the quasi-reversibility, and we specify the form of the stationary measures. We revisit two canonical examples of quasi-reversible queueing systems, Whittle networks and order-independent queues. Lastly, we focus on the problem of admission control and leverage our results in the frameworks of optimization and reinforcement learning.

Reconstructing facial images from black-box recognition models poses a significant privacy threat. While many methods require access to embeddings, we address the more challenging scenario of model inversion using only similarity scores. This paper introduces DarkerBB, a novel approach that reconstructs color faces by performing zero-order optimization within a PCA-derived eigenface space. Despite this highly limited information, experiments on LFW, AgeDB-30, and CFP-FP benchmarks demonstrate that DarkerBB achieves state-of-the-art verification accuracies in the similarity-only setting, with competitive query efficiency.

E-mail: nishio@ict.eng.isct.ac.jp Abstract --This paper proposes Load-aware Tram-FL, an extension of Tram-FL that introduces a training scheduling mechanism to minimize total training time in decentralized federated learning by accounting for both computational and communication loads. The scheduling problem is formulated as a global optimization task, which--though intractable in its original form--is made solvable by decomposing it into node-wise subproblems. T o promote balanced data utilization under non-IID distributions, a variance constraint is introduced, while the overall training latency, including both computation and communication costs, is minimized through the objective function. Simulation results on MNIST and CIF AR-10 demonstrate that Load-aware Tram-FL significantly reduces training time and accelerates convergence compared to baseline methods. Federated learning (FL) enables model training without exporting data, making it particularly effective for privacy-sensitive applications.

The performance of an optimizer on large-scale deep learning models depends critically on fine-tuning the learning rate, often requiring an extensive grid search over base learning rates, schedules, and other hyperparameters. In this paper, we propose a principled framework called GALA (Gradient Alignment-based Learning rate Adaptation), which dynamically adjusts the learning rate by tracking the alignment between consecutive gradients and using a local curvature estimate. Guided by the convergence analysis, we formulate the problem of selecting the learning rate as a one-dimensional online learning problem. When paired with an online learning algorithm such as Follow-the-Regularized-Leader, our method produces a flexible, adaptive learning rate schedule that tends to increase when consecutive gradients are aligned and decrease otherwise. We establish a data-adaptive convergence rate for normalized SGD equipped with GALA in the smooth, nonconvex setting. Empirically, common optimizers such as SGD and Adam, when augmented with GALA, demonstrate robust performance across a wide range of initial learning rates and perform competitively without the need for tuning.

Multi-objective optimization (MOO) is a critical area of research in the field of optimization, focusing on problems that involve multiple conflicting objectives. Unlike traditional single-objective optimization, where the goal is to find a single optimal solution, MOO seeks to find a set of solutions that represent the trade-offs between the different objectives. MOO has been successfully applied across diverse fields, from energy system design Marler and Arora (2004) to healthcare treatment planning Craft et al. (2006), demonstrating its versatility in balancing conflicting objectives. In machine learning, multi-objective problems play a significant role in various applications, including recommender systems Milojkovic et al. (2019); Jannach (2022); Zaizi et al. (2023), where they help balance multiple conflicting criteria, and multi-task learning Sener and Koltun (2018); Crawshaw (2020), where they optimize performance across multiple related tasks simultaneously. Furthermore, addressing multi-objective problems in real-world scenarios can be computationally demanding Lin et al. (2022), particularly in high-dimensional settings Nguyen and Tran (2024), as it involves assessing multiple conflicting objectives, which increases computational costs.

In this paper, we address the problem of identifying the Pareto Set under feasibility constraints in a multivariate bandit setting. Specifically, given a $K$-armed bandit with unknown means $μ_1, \dots, μ_K \in \mathbb{R}^d$, the goal is to identify the set of arms whose mean is not uniformly worse than that of another arm (i.e., not smaller for all objectives), while satisfying some known set of linear constraints, expressing, for example, some minimal performance on each objective. Our focus lies in fixed-confidence identification, for which we introduce an algorithm that significantly outperforms racing-like algorithms and the intuitive two-stage approach that first identifies feasible arms and then their Pareto Set. We further prove an information-theoretic lower bound on the sample complexity of any algorithm for constrained Pareto Set identification, showing that the sample complexity of our approach is near-optimal. Our theoretical results are supported by an extensive empirical evaluation on a series of benchmarks.

This book offers a hands-on introduction to building and understanding federated learning (FL) systems. FL enables multiple devices -- such as smartphones, sensors, or local computers -- to collaboratively train machine learning (ML) models, while keeping their data private and local. It is a powerful solution when data cannot or should not be centralized due to privacy, regulatory, or technical reasons. The book is designed for students, engineers, and researchers who want to learn how to design scalable, privacy preserving FL systems. Our main focus is on personalization: enabling each device to train its own model while still benefiting from collaboration with relevant devices. This is achieved by leveraging similarities between (the learning tasks associated with) devices that are encoded by the weighted edges (or links) of a federated learning network (FL network). The key idea is to represent real-world FL systems as networks of devices, where nodes correspond to device and edges represent communication links and data similarities between them. The training of personalized models for these devices can be naturally framed as a distributed optimization problem. This optimization problem is referred to as generalized total variation minimization (GTVMin) and ensures that devices with similar learning tasks learn similar model parameters. Our approach is both mathematically principled and practically motivated. While we introduce some advanced ideas from optimization theory and graph-based learning, we aim to keep the book accessible. Readers are guided through the core ideas step by step, with intuitive explanations.

MOMAV (Marco's Omnidirectional Micro Aerial Vehicle) is a multirotor drone that is fully actuated, meaning it can control its orientation independently of its position. MOMAV is also highly symmetrical, making its flight efficiency largely unaffected by its current orientation. These characteristics are achieved by a novel drone design where six rotor arms align with the vertices of an octahedron, and where each arm can actively rotate along its long axis. Various standout features of MOMAV are presented: The high flight efficiency compared to arm configuration of other fully-actuated drones, the design of an original rotating arm assembly featuring slip-rings used to enable continuous arm rotation, and a novel control allocation algorithm based on sequential quadratic programming (SQP) used to calculate throttle and arm-angle setpoints in flight. Flight tests have shown that MOMAV is able to achieve remarkably low mean position/orientation errors of 6.6mm, 2.1° (σ: 3.0mm, 1.0°) when sweeping position setpoints, and 11.8mm, 3.3° (σ: 8.6mm, 2.0°) when sweeping orientation setpoints.

Parameter tuning for vehicle controllers remains a costly and time-intensive challenge in automotive development. Traditional approaches rely on extensive real-world testing, making the process inefficient. We propose a multi-fidelity Bayesian optimization approach that efficiently learns optimal controller parameters by leveraging both low-fidelity simulation data and a very limited number of real-world experiments. Our approach significantly reduces the need for manual tuning and expensive field testing while maintaining the standard two-stage development workflow used in industry. The core contribution is the integration of an auto-regressive multi-fidelity Gaussian process model into Bayesian optimization, enabling knowledge transfer between different fidelity levels without requiring additional low-fidelity evaluations during real-world testing. We validate our approach through both simulation studies and realworld experiments. The results demonstrate that our method achieves high-quality controller performance with only very few real-world experiments, highlighting its potential as a practical and scalable solution for intelligent vehicle control tuning in industrial applications.

This paper presents a periodic bipedal gait learning method using reward composition, integrated with a real-time gait planner for humanoid robots. First, we introduce a novel gait planner that incorporates dynamics to design the desired joint trajectory. In the gait design process, the 3D robot model is decoupled into two 2D models, which are then approximated as hybrid inverted pendulums (H-LIP) for trajectory planning. The gait planner operates in parallel in real time within the robot's learning environment. Second, based on this gait planner, we design three effective reward functions within a reinforcement learning framework, forming a reward composition to achieve periodic bipedal gait. This reward composition reduces the robot's learning time and enhances locomotion performance. Finally, a gait design example and performance comparison are presented to demonstrate the effectiveness of the proposed method.