Hyperparameter search: grid, random, elimination
With the same budget a grid search finds 0.33 while a random search finds 0.83. The reason is not luck but geometry.
4 steps
170 XP
A free account is needed
Start the lesson →
Sources
- Bergstra, J. & Bengio, Y. 2012 · Random Search for Hyper-Parameter Optimization · JMLR, 13, 281-305
- Li, L. et al. 2018 · Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimization · JMLR, 18(185)
- Hutter, F., Kotthoff, L. & Vanschoren, J. 2019 · Automated Machine Learning: Methods, Systems, Challenges · Springer
ML Academy · an interactive machine learning course that runs in your browser ·
All lessons