Create a model whose training part is random forest and its experimental part is binary classificat
Create a model whose training part is random forest and its experimental part is binary classification using cross-validation
Hello friends
I want to implement the model inside the article I attached with Rapid Miner.But I encountered the following problems:
1- How can I create a model using cross-validation to use random forest in the experimental section and binary classification in the training section? (90% training and 10% experiment)
2. How do I do the Pearson correlation coefficient in Ripper Miner?
3- How to implement the diagram ROC for the desired model?
Please help
General structure of the model:
Find more posts tagged with
Sort by:
1 - 2 of
21
thank you
Sort by:
1 - 1 of
11
You may want to watch some of the training videos which will help you out with using RapidMiner Studio in general.
Here are some pointers on the questions you have asked:
1) This question is a bit unclear, as Random Forest IS a binary classification algorithm. So when you use it inside Cross-Validation it will be used on the training set and applied on the testing set. You don't use two different algorithms in that setup.
2) There is an operator to calculate Pearson coefficient called "Correlate" or you can use "Weight by Correlation" if you want to look at correlation with your target variable (label in RapidMiner).
3) The ROC is one of the outputs of the "Performance(Binary Classification)" operator. You should watch a few of the tutorial videos on setting up a modeling workflow to see more details.
Here are some pointers on the questions you have asked:
1) This question is a bit unclear, as Random Forest IS a binary classification algorithm. So when you use it inside Cross-Validation it will be used on the training set and applied on the testing set. You don't use two different algorithms in that setup.
2) There is an operator to calculate Pearson coefficient called "Correlate" or you can use "Weight by Correlation" if you want to look at correlation with your target variable (label in RapidMiner).
3) The ROC is one of the outputs of the "Performance(Binary Classification)" operator. You should watch a few of the tutorial videos on setting up a modeling workflow to see more details.
Here are some pointers on the questions you have asked:
1) This question is a bit unclear, as Random Forest IS a binary classification algorithm. So when you use it inside Cross-Validation it will be used on the training set and applied on the testing set. You don't use two different algorithms in that setup.
2) There is an operator to calculate Pearson coefficient called "Correlate" or you can use "Weight by Correlation" if you want to look at correlation with your target variable (label in RapidMiner).
3) The ROC is one of the outputs of the "Performance(Binary Classification)" operator. You should watch a few of the tutorial videos on setting up a modeling workflow to see more details.