Learning Platform User Subscription Renewal Predictions AI Hackathon

음..사용해도 되지 않나요? train 데이터에 대한 feature engineering을 한 경우, test_data에 대해서 적용할 때는 똑같이 데이터 전처리만 동일하게 하고 적용해도 되는거 아닌가 해서요..! 

예를 들어드리면 0,1,2,3,4,5를 갖는 범주형 데이터가 있는데 train data에 0,1,2,3,5의 경우만 존재해서 get_dummies()를 쓰면 4가 반영되지 않습니다. 그런데 만약 test data에는 0,1,2,3,4,5가 전부 존재한다면 input data shape이(데이터의 열 개수) train data의 경우와 test data의 경우가 다르게 됩니다