Genetic algorithms can be used for optimizing parameters of a graph partition and allow informed problem-specific mutation functions. NP-hard problems like 3SAT provide opportunity for optimizing genetic algorithms by generating random problems which can be used for feature discovery of graph partitioning solutions. Using this model, the probability of hardness of 3SAT can be defined by selecting problems for phase transition topography. This project aims to use Intel Math Kernel Library for Python and Numpy implementations benchmarked against native Python and Numpy for performance comparison.
- Projects 0
- Followers 1