SimonBlanke /
Gradient-Free-Optimizers
| Total Complexity | 4 |
| Total Lines | 30 |
| Duplicated Lines | 0 % |
| Changes | 0 | ||
| 1 | # Author: Simon Blanke |
||
| 2 | # Email: [email protected] |
||
| 3 | # License: MIT License |
||
| 4 | |||
| 5 | |||
| 6 | import numpy as np |
||
| 7 | |||
| 8 | from ..local import StochasticHillClimbingOptimizer |
||
| 9 | from ...search import Search |
||
| 10 | |||
| 11 | |||
| 12 | class SimulatedAnnealingOptimizer(StochasticHillClimbingOptimizer, Search): |
||
| 13 | def __init__(self, search_space, annealing_rate=0.99, start_temp=100): |
||
| 14 | super().__init__(search_space) |
||
| 15 | self.annealing_rate = annealing_rate |
||
| 16 | self.temp = start_temp |
||
| 17 | |||
| 18 | # use _consider from StochasticHillClimbingOptimizer |
||
| 19 | |||
| 20 | def _accept_default(self): |
||
| 21 | return np.exp(-self._score_norm_default() / self.temp) |
||
| 22 | |||
| 23 | def _accept_adapt(self): |
||
| 24 | return self._score_norm_adapt() * self.temp |
||
| 25 | |||
| 26 | def evaluate(self, score_new): |
||
| 27 | super().evaluate(score_new) |
||
| 28 | |||
| 29 | self.temp = self.temp * self.annealing_rate |
||
| 30 |
