Text Classification Problem : Wrote a module to classify Amazon-Product Reviews as favourable/unfavourable. Achieved accuracy of 78% and an F1 score of .81 using Logistic Regression on a test-train split of 20%, where total records were around 50000.