ankushtale / Bayesian_Global_Optimization_in_CNN

<Work in progress>This project demonstrates application of Gaussian Processes to effectively selecting best hyper-parameter for a model. This is part of CSE569-Fundamentals of Statistical Learning Course by Prof. Hemanth Venkateswara.
0 stars 1 forks source link

Grid Search vs Bayesian Optimization #1

Open Kummerming opened 4 years ago

Kummerming commented 4 years ago

Hi, I am also trying to apply the Bayesian Optimization in CNN. I have looked through your codes. It seems that you use Grid Search for hyperparameters. I am not sure whether I understand right. Would you mind having a further discussion? Best, Ming

ankushtale commented 4 years ago

Hey Ming,

Apologize, we haven't updated the code since the last time. Let me push the latest update. But if you want to just implement Bayesian Optimization, I suggest you try out BayesOpt https://github.com/fmfn/BayesianOptimization

ankushtale commented 4 years ago

Have updated the codebase

Kummerming commented 4 years ago

Have updated the codebase

Thank you Ankush for the recommendation and the update.