jieguangzhou commented on issue #9725:
URL: 
https://github.com/apache/dolphinscheduler/issues/9725#issuecomment-1108445809

   I add the GridSearch feature to the example like this. That we can search 
for the best parameters of a model.
   
   <img width="491" alt="image" 
src="https://user-images.githubusercontent.com/31528124/165078096-54543d05-36db-425f-b988-745af105db03.png";>
   
   After the model was trained, we also can see the parameter search report on 
the MLflow dashboard like this
   
   <img width="2317" alt="image" 
src="https://user-images.githubusercontent.com/31528124/165079578-d23a2098-8df3-44e0-9863-fb73499f79c0.png";>
   
   
   The  value of key "search_params" above is "max_depth=[5, 
10];n_estimators=[100, 200]" for lightgbm.
   
   We also can search more params, all of the params we can see In API of  the 
algorithm, for example:
   svm: "kernel=['linear', 'poly', 'rbf'];C=[0.5, 1.0]" ;
   lr: "penalty=['l1', 'l2'];C=[0.5, 1.0]" ;
   xgboost: max_depth=[5, 10];n_estimators=[100, 200]
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to