Due to recent updates, all users are required to create an Altair One account to login to the RapidMiner community. Click the Register button to create your account using the same email that you have previously used to login to the RapidMiner community. This will ensure that any previously created content will be synced to your Altair One account. Once you login, you will be asked to provide a username that identifies you to other Community users. Email us at Community with questions.
[S] Can I avoid Optimize Parameters algorithms' subparameters incorrect testing?
Contributor II
For sure the title does not explain itself, I tried!
The problem is very simple, for example, if I set a k-NN with a grid search for:
k: 1-100 (10 steps).
measure_types: mixed, numerical and bregman.
numerical measure: 10 of them.
The process executes 10 x 3 x 10 (300) instead of 10 x 2 + 10 x 10 (120), because it tries out the 10 different numerical measures for mixed and bregman measure types, even though they do not apply to them.
Is there a way to avoid this without programming X different optimization nodes, where X is the number of instances this error occurs? X can get quite large, and the 300/120 ratio even larger!
If there is not, any insight on how to better approach my quite-brute solution space search? ;D
Thanks in advance!!
The problem is very simple, for example, if I set a k-NN with a grid search for:
k: 1-100 (10 steps).
measure_types: mixed, numerical and bregman.
numerical measure: 10 of them.
The process executes 10 x 3 x 10 (300) instead of 10 x 2 + 10 x 10 (120), because it tries out the 10 different numerical measures for mixed and bregman measure types, even though they do not apply to them.
Is there a way to avoid this without programming X different optimization nodes, where X is the number of instances this error occurs? X can get quite large, and the 300/120 ratio even larger!
If there is not, any insight on how to better approach my quite-brute solution space search? ;D
Thanks in advance!!
0
Answers
the easiest way would be to add two optimization operators, one for each type of measures.
Best regards,
Marius