Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
LeiZhang 37cdbd5975 | 1 year ago | |
---|---|---|
.. | ||
README.md | 1 year ago | |
rosenbrock_LaMCTS.py | 1 year ago |
Latent Action Monte Carlo Tree Search (LA-MCTS)
Since LaNAS works very well on NAS datasets, e.g. NASBench-101, and the core of the algorithm can be easily generalized to other problems, we extend it to be a generic solver for black-box function optimization. LA-MCTS further improves by using a nonlinear classifier at each decision node in MCTS and use a surrogate (e.g., a function approximator) to evaluate each sample in the leaf node. The surrogate can come from any existing Black-box optimizer (e.g., Bayesian Optimization). The details of LA-MCTS can be found in the following paper.
E.g. PYTHONPATH='./' python examples/LaMCTS/rosenbrock_LaMCTS.py
Modify the following section of comparison/xbbo_benchmark.py
:
test_algs = ["lamcts"]
And run PYTHONPATH='./' python comparison/xbbo_benchmark.py
in the command line.
Method | Minimum | Best minimum | Mean f_calls to min | Std f_calls to min | Fastest f_calls to min |
---|---|---|---|---|---|
XBBO(lamcts) | 0.440+/-0.040 | 0.407 | 81.7 | 39.666 | 4 |
超参搜索(黑盒优化)框架
Python other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》