search

eulerfan / Russian_twitter_trolls

0 stars 0 forks source link

Cross Validation #12

Open mzakariaCERN opened 5 years ago

mzakariaCERN commented 5 years ago

One more thing you need to work on... very important! You need to do cross validation. That is, you need to repeat the same analysis multiple time (with different data each time) to make sure that we didn't just get lucky with the sample if data we used in training/testing. Take a look at Lantz P:340

eulerfan commented 5 years ago

I made an effort to do the CVs for the other models, but I only had time for the decision tree model. I will definitely go back later and update this part.