Rethinking the parallelization of random-restart hill climbing: a case study in optimizing a 2-opt TSP solver for GPU execution.
Molly A. O'NeilMartin BurtscherPublished in: GPGPU@PPoPP (2015)
Keyphrases
- hill climbing
- traveling salesman problem
- search space
- sat solvers
- clause learning sat solvers
- simulated annealing
- parallel processing
- parallel execution
- combinatorial optimization
- search algorithm
- ant colony optimization
- max min
- real time
- genetic algorithm ga
- metaheuristic
- search procedure
- direct search
- parallel computation
- steepest ascent
- systematic search
- search strategy
- parallel computing
- genetic algorithm
- heuristic search
- hill climbing algorithm
- optimal solution
- hybrid algorithms
- graphics hardware
- heuristic function
- beam search
- path finding
- rule learning
- case study
- random walk
- constraint satisfaction problems
- tabu search
- constraint satisfaction
- branch and bound
- gpu implementation
- graphics processors
- neural network
- greedy search
- parallel implementation
- search tree
- optimization problems
- lower bound