Safe Learning for Near Optimal Scheduling

05/19/2020

∙

In this paper, we investigate the combination of synthesis techniques and learning techniques to obtain safe and near optimal schedulers for a preemptible task scheduling problem. We study both model-based learning techniques with PAC guarantees and model-free learning techniques based on shielded deep Q-learning. The new learning algorithms have been implemented to conduct experimental evaluations.

READ FULL TEXT

Safe Learning for Near Optimal Scheduling

Directed Exploration in PAC Model-Free Reinforcement Learning

Scheduling for Urban Air Mobility using Safe Learning

Near Optimal Task Graph Scheduling with Priced Timed Automata and Priced Timed Markov Decision Processes

Learning to search efficiently for causally near-optimal treatments

Near-Optimal Cryptographic Hardness of Agnostically Learning Halfspaces and ReLU Regression under Gaussian Marginals

Simple and near-optimal algorithms for hidden stratification and multi-group learning

CycleSense: Detecting Near Miss Incidents in Bicycle Traffic from Mobile Motion Sensors

Safe Learning for Near Optimal Scheduling

Related Research

Directed Exploration in PAC Model-Free Reinforcement Learning

Scheduling for Urban Air Mobility using Safe Learning

Near Optimal Task Graph Scheduling with Priced Timed Automata and Priced Timed Markov Decision Processes

Learning to search efficiently for causally near-optimal treatments

Near-Optimal Cryptographic Hardness of Agnostically Learning Halfspaces and ReLU Regression under Gaussian Marginals

Simple and near-optimal algorithms for hidden stratification and multi-group learning

CycleSense: Detecting Near Miss Incidents in Bicycle Traffic from Mobile Motion Sensors