Quora_QuestionPair

Employed Natural Language Processing techniques such as removal of Stopwords, Punctuations and Hyperlinks to prepare the Dataset(consisting 404290 rows) and also applied techniques such as Tokenization and Stemming • Extracted Basic features and Advance Features consisting of Fuzz features and explored the features importance • Transformed the texts to numerical vectors using TF-IDF Vectorizer and fitted Logistic Regression and Xgboost Model and did Hyperparameter Tuning on Xgboost model to get Auc score of 0.91 and accuracy 83%