Optimistic Learning with Applications to Caching Networks
Ph.D. thesis on optimistic online learning with applications to caching networks.
On the Dynamic Regret of Following the Regularized Leader: Optimism with History Pruning
Dynamic-regret analysis of FTRL with optimistic predictions and a history-pruning technique that improves bounds.
Optimistic Learning for Communication Networks
Monograph on optimistic online learning for communication networks (caching, routing, resource management).
george-iosifidis
Online Caching with no Regret: Optimistic Learning via Recommendations
An optimistic learning framework for online caching that leverages recommendations to improve regret without sacrificing worst-case guarantees.
Optimistic Online Non-stochastic Control via FTRL
Optimistic FTRL controllers for online non-stochastic control with improved performance under accurate predictions.
Optimistic No-regret Algorithms for Discrete Caching
No-regret algorithms for discrete caching that integrate untrusted predictions for improved performance.
Online Caching with Optimistic Learning
Online caching with optimistic learning, exploiting predictions while keeping no-regret guarantees.