Building a basic block instruction scheduler with reinforcement learning and rollouts

Machine Learning

Volumn 49, Issue 2-3, 2002, Pages 141-160

  McGovern, Amy ^a   Moss, Eliot ^a   Barto, Andrew G ^a  

^a Biologically Inspired Neural and Dynamical Systems Laboratory   (United States)

Author keywords

Instruction scheduling; Reinforcement learning; Rollouts

Indexed keywords

BENCHMARKING; BLOCK CODES; COMPUTER AIDED INSTRUCTION; COMPUTER ARCHITECTURE; COMPUTER SIMULATION; HEURISTIC PROGRAMMING; LEARNING ALGORITHMS;

COMPAQ ALPHA MACHINES; HEURISTIC ALGORITHMS; HEURISTIC SCHEDULER; INSTRUCTION SCHEDULING; REINFORCEMENT LEARNING SCHEDULER; ROLLOUT LEARNING;

LEARNING SYSTEMS;

EID: 0036832961     PISSN: 08856125     EISSN: None     Source Type: Journal    
DOI: 10.1023/A:1017976211990     Document Type: Article

References (26)

1. Expected-outcome: A general model of static evaluation
2. Differential training of rollout policies
3. Rollout algorithms for combinatorial optimization
4. Learning and improving backgammon strategy
5. Advantage updating applied to a differential game
6. Scheduling straight-line code using reinforcement learning and rollouts
7. Basic-block instruction scheduling using reinforcement learning and rollouts
8. Learning to schedule straight-line code
9. Least-cost instruction selection in DAGs in NP-complete
10. SPEC describes SPEC95 products and benchmarks
11. Learning internal representations by error propagation
12. Applying reinforcement learning to instruction scheduling within basic blocks
13. ATOM: A system for building customized program analysis tools
14. The character of the instruction scheduling problem
15. Learning to predict by the method of temporal differences
16. On-line policy improvement using Monte-Carlo search
17. Decision tree induction based on efficient tree restructuring
18. Two kinds of training information for evaluation function learning
19. Constructive function approximation
20. Relative value function approximation
21. Approximate dynamic programming for real-time control and neural modeling
22. Rollouts
23. A reinforcement learning approach to job-shop scheduling