Random Forest Regression Analysis - Comprehension problem

Question

I have an deeply understanding-problem with Random Forest Regression. Target is a university project: We have to do a random forest regression analytics with financial data in R. I already read many hours in random forest examples, most of them are classification type like to predict a stock value is go up or down. In case of regression I stand on the line .. My thoughts are the following:

If I had a data set like the following structure:

Date | Open | High | Low | Close | Volume

...I could add some technical instruments as RSI, SMA and so on

Then I split the data set to training and test data, execute the random forest procedure and close this with predicting on the test data. But is this really the intent of a regression analytics with random forest? I guess a 'correct' regression analysis is to compare two stocks and see if they correlate to predict the value of one stock based on the other stock - but on the other side this is a classical regression analysis - in the absence of Random Forest Algorithm. I have a really big problem with understanding the purpose ...

Answer 1

In case of the actual regression procedure, you could do a train/test split or either K-fold or any other cross val methodology to evaluate the performance of the model. If your satisfied with the performance, you could then use ALL your data for training. To predict the future, in sklearn you can use the predict(X) function.