question-please

afsaneafsane Member Posts: 4 Learner I
Hello
How can I create a model that uses random forest in training and binary classification in experiments?
90% for training
10% for testing
And use cross validation
Please help !!
And how can I get FPR TPR FNR TNR values?
I went to the contents, but I didn't get an answer. We can help


Best Answer

  • varunm1varunm1 Moderator, Member Posts: 1,207 Unicorn
    edited April 2020 Solution Accepted
    Hello @afsane

    I recommend you go to academy.rapidminer.com to check tutorials on how to build models and using validations. In the meantime you can also check a sample process attached in this thread. It is based on a sample dataset called Titanic, I split that into 90% Train and 10% test. I cross-validated 90% train data using decision tree (you can use any algorithm) and applied the model on 10% test data. To check this in your rapidminer, just download it and go to rapidminer studio then go to FILE --> Import Process and point to this file. You can see the process.

    TPR and TNR are sensitivity and specificity that are available in performance (Binominal) operator. You can calculate the rest based on the confusion matrix.
    Regards,
    Varun
    https://www.varunmandalapu.com/

    Be Safe. Follow precautions and Maintain Social Distancing

Answers

  • afsaneafsane Member Posts: 4 Learner I
    edited April 2020

    Hello
    Thank you for answering me because some of the things I didn't know were solved with your guidance.
    I searched a lot for the following, but to no avail.
    I want to implement the method in the article with Rapid Miner.
    However, I do not know of any of the following:
    1. I don't know where to add the binary classification for the test section.
    2. I don't know where to set 90% for training and 10% for testing.
    3- I don't know where to calculate Pearson'sThe correlation coefficient

    I have added the article in question in this post. Please read
    You can help me if you know
    Please






Sign In or Register to comment.