Altair RISE

A program to recognize and reward our most engaged community members

Nominate Yourself Now!

how to use optimized operator in cross project.

how to use optimized operator in cross project. Suppose we have one training data and five testing data, how to use optimized operator and get the value of k and optimal value for each testing data set.

Find more posts tagged with

AI Studio

Accepted answers

All comments

lionelderkrikor

Hi @Ma-123,

I have difficulties to understand what you want to do...
Can you explain in more detail what you want to do by giving what you have and what you want to obtain.
Maybe can you share your process and your data...

Regards,

Lionel

Ma-123

lionelderkrikor ,
Thanks for your reply. I have designed cross project in Rapidminer with KNN technique.
I have seven dataset. I want to do one data set as training and 6 datasets as setting. So there is 42 combinations. I have built model for that. But i want to use optimized operator. Problem is that when i use optimized operator, only training data set is optimized. So i want to optimized all the testing datasets and want to get optimal value of k of each testing datasets.
If you wish i can share my process.

Regards

varunm1

Hello @Ma-123

Problem is that when I use optimized operator, only training data set is optimized.

Yes, optimization, hyperparameter tuning, feature selection are applied to training data. If you want to use your test dataset as the performance reference to tune your model hyperparameters (such as K value in KNN) then it is no longer a test dataset (it can be treated as validation).

Now, coming to your question about tuning hyperparameter K value based on your 6 datasets, I can think of a manual way where you will set a macro that will help set different "k" value at different runs. I attached a process (Main_Iteration_K_Using_Loop) with clear comments associated with each operator. You can download and import this process into your rapidminer studio and run it to see its working. The "Log" operator output will provide you the "K" value and its relevant performance on test dataset.

One downside is it will go through all K values in the increments of 1. There are other ways to use macros. This can help you for now I guess. If I get something in my mind, I will inform you here.

@lionelderkrikor if you get any other thoughts based on this process please let me know.

Let us know if you have any questions.

Main_Iteration_K_Using_Loop.rmp

Ma-123

@varunm1
Thanks for reply
please send me the .properties file

Ma-123

Hi @ varunm1
The file you attached are .rmp file. I need .properties file to open it into rapid miner.
Thanks and regards
Madhu

varunm1

@Ma-123

Please find attached zip file with .rmp and .properties files. You need to paste these in the repository of the .rapidminer folder. You can also import the .rmp file using FILE --> Import Process in Rapidminer.

Prop_Files.zip

Ma-123

Hi @Varun
Thanks for sharing your operator. But in this operator there is no optimized operator. To clear undertaing of my problem i m sending you my operator and my data set at your email id.

varunm1

Hello @Ma-123

You can send me in Private Message on the community messaging. The earlier solution is without optimize parameters. In the meantime, you can take a look at the new option I created. This is a scalable method and can be used with any operator.

Please find attached .rmp file, you don't need properties file. You just need to go to rapidminer software and click on FILE --> Import process and point to the location where the .rmp file in this thread is downloaded on your PC.

Ma-123

Hello varunm1
How can send u private message on the community message. Plz let me know.

nabilaftha

hello can you tell me how to do cross project for 4 dataset? I want to do 1 dataset as testing and 3 datasets as training.