Assignment 1 - Analysing tweets with hashtag 'shutdown' using feature extraction and decomposition from scikit learn library. Data is provided in a file. We identified most re-tweeting accounts and most co-occurring hashtags by using different vectorizations.
Assignement 2 - Sentiment analysis on airline review data using feature extraction, model selection, linear model, and naive bayes from scikit learn library. We trained two prediction models, Logistic regression and Multinomial Naive Bayes using all the sentiment data and could predict sentiment in about 76%.
Assignment 3 - Clustering of song lyrics using feature extraction, cluster and decomposition from scikit learn library. Data is provided in a file. We identified five clusters using elbow method and five songs to each centroid.