Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 155 for “"Convergence Guarantees"”.
-
PER PLAYER LOCAL EXTRAGRADIENT IN MULTIPLAYER FEDERATED LEARNING: CONVERGENCE GUARANTEES AND PRACTICAL PERFORMANCE
… with the extragradient method. We establish convergence guarantees for PEARL-EG for both strongly monotone and monotone problems, explain its benefits over prior approaches, and present numerical experiments that complement the theory by exploring practical performance and limitations.
-
Fixed-Time Reinforcement Learning-based Control for Safe Autonomy
… safe learning architectures lack finite time convergence guarantees, rendering these algorithms impractical for real-world applications. In this dissertation, we enable safe autonomy by endowing autonomous systems with safety-critical control frameworks predicated on online reinforcement …
-
FINITE-TIME CONVERGENCE ANALYSIS OF ACTOR-CRITIC ALGORITHMS
… In the thesis, we provide the first global convergence guarantees for two-timescale actor-critic (AC) on linear quadratic regulator (LQR) problems, achieving epsilon-optimality with provable sample complexity. These results are generalized to Markov Decision Processes (MDPs) and extended to …
-
Optimisation and computational optimal control via local pursuit
… to noise-free non-stationary ones, preserving convergence guarantees and matching the quality of the ‘full-knowledge’ benchmark approach at substantially lower computational cost. Subsequently, we account for noise in both agent and obstacle observations, where convergence guarantees are …
-
Reinforcement learning for multi-agent and robust control systems
… for these settings, supported by theoretical convergence guarantees. In setting i, a team of collaborative MARL agents is connected via a communication network, without the coordination of any central controller. With only neighbor-to-neighbor communications, we introduce decentralized …
-
Extragradient-type Methods for (L0, L1)-Lipschitz Variational Inequality Problems
… algorithms. Traditionally, the convergence of these algorithms has been analyzed under the L-Lipschitz assumption, but recent research has found that this assumption is often overly restrictive in practice and does not accurately capture the operators encountered in many modern …
-
Factoring over probability simplices : from theory to applications
… we establish subsequential and full-sequence convergence guarantees. This analysis framework proves remarkably versatile: we successfully apply it to both our probabilistic tensor factorization models and to an adaptive graph-regularized denoising method for spatial transcriptomics, where the …
-
Regret minimisation and system-efficiency in route choice
… objective a moving target. As a side effect, no convergence guarantees exist for the general MARL setting. This thesis exploits a particular MARL problem, namely route choice (where selfish drivers aim at choosing routes that minimise their travel costs), to deliver convergence guarantees. We are …
-
Dynamic Algorithms and Asymptotic Theory for Lp-norm Data Analysis
… (SVD), and asymptotic convergence guarantees. However, PCA has been shown to be highly sensitive to faulty measurements (outliers) because of its reliance on the outlier-sensitive L2-norm. Arguably, the most straightforward approach to impart robustness against outliers …
-
Faster Conditional Gradient Algorithms for Machine Learning
… either locally or globally improve upon existing convergence guarantees. In Chapters 2-4 we focus on the case where the objective function is smooth and strongly convex and the feasible region is a polytope, and in Chapter 5 we focus on the case where the function is generalized self-concordant …
-
Nonconvex Optimization Methods with Applications to Portfolio Selection and Hybrid Systems
… and develop efficient algorithms with standard convergence guarantees. The effectiveness of the algorithms is demonstrated in detail for both applications with numerical experiments.
-
Algorithms and Theory for Saddle Point Problems with Relaxed Structural Conditions
… show that our proposed method achieves a linear convergence rate under either condition. These results extend linear convergence guarantees beyond the strong convexity--strong concavity setting typically required in existing analyses. The second research chapter of this dissertation studies …
-
GRADIENT DESCENT METHODS IN MODERN MACHINE LEARNING PROBLEMS: PROVABLE GUARANTEES
… is non-convex and non-smooth, the standard convergence guarantees of GD for convex loss functions cannot be applied. Instead, through a kernel perspective, we demonstrate that when fresh data arrives in a stream, SGD ensures the exponential convergence of the average prediction error. In the …
-
Projection-Free and Accelerated Methods for Constrained Optimization and Saddle-Points Problems
… offering robust solutions with theoretical guarantees. In the first part of the dissertation, we propose an inexact-proximal accelerated gradient method to address nonconvex stochastic composite optimization problems, where the objective function is composed of smooth and nonsmooth …
-
BUILDING AND MAINTAINING STRUCTURE IN DISTRIBUTED NETWORKS
… a tight analysis of tolerable churn and recovery guarantees against catastrophic failures like denial-of-service attacks. Finally, the thesis explores self-stabilizing overlay networks through graph linearization augmented with untrusted advice. This mechanism achieves optimal self-stabilization …
-
Efficient Hierarchical Global Motion Planning for Autonomous Vehicles
… non-trivial, including issues such as algorithm convergence, its correspondence to system controllability, optimality of solution, computational complexity, and dynamic environments. In response to these challenges, a hierarchical algorithm will be introduced that provides (a) decreased …
-
Across the Reinforcement Learning and Large Language Model Spectrum
… The first part develops rigorous theoretical guarantees for RL on tabular Markov decision processes (MDPs), specifically, variance-dependent regrets for MDPs and latent MDPs, and a fundamental progress on completely horizon-free regrets. The second part studies how language models retrieve …
-
SOME PRIMAL-DUAL THEORY FOR SUBGRADIENT METHODS FOR STRONGLY CONVEX OPTIMIZATION
… method. These equivalences enable optimal convergence guarantees in terms of both their classic primal gap and a not previously analyzed dual gap for strongly convex optimization. Consequently, our theory provides these classic methods with simple, optimal stopping criteria and optimality …
Page 1 of 8