Justin Boyan

Tags

Active Learning, Applications, Markov Decision Processes, Optimization, Reinforcement Learning

Papers

• Value Function Based Production Scheduling (1998)
  Production scheduling in which we account for a probability distribution on future jobs by means of kernel-based value function approximation
• Q2: Memory-based active learning for optimizing noisy continuous functions (1998)
  Maximizing a very noisy function in k-dimensional space with few samples
• Learning Evaluation Functions for Global Optimization and Boolean Satisfiability (1998)
  
• Using Prediction to Improve Combinatorial Optimization Search (1997)
  Automatically improving combinatorial search by reinforcement-learning-style analysis of earlier runs
• Algorithms for Approximating Optimal Value Functions in Acyclic Domains (1996)
  Using "Rollouts" to make value-function-based RL more practical
• Generalization in Reinforcement Learning: Safely Approximating the Value Function (1995)
  An introduction to the ways that naive application of function approximation of value functions can fail.
• Proceedings of the Workshop on Value Function Approximation, Machine Learning Conference 1995. (1995)
  Short talks from a workshop about value function approximation