Used gradient boosting regression to predict Rotten Tomatoes scores for movies of various genres. Dataset retrieved from Kaggle, and categorical encoding was done on my own. Will update this soon, but some quick facts: mean absolute error for predicting Tomatometer score was about 19. I was able to reduce it further by increasing the number of estimators, but the training time did take much longer. Films rated NR showed the least mean absolute error, about 12.
kaisubr / rotten-tomatoes-prediction Goto Github PK
View Code? Open in Web Editor NEWUsed gradient boosting regression to predict Rotten Tomatoes scores for movies of various genres.
License: Apache License 2.0