Twitter Sentiment Analysis using Deep Learning

In this report, address the problem of sentiment classification on twitter dataset. used a number of<br>machine learning and deep learning methods to perform sentiment analysis. In the end, used a majority<br>vote ensemble method with 5 of our best models to achieve the classification accuracy of 83.58% on<br>kaggle public leaderboard. compared various different methods for sentiment analysis on tweets (a<br>binary classification problem). The training dataset is expected to be a CSV file of type tweet_id,<br>sentiment, tweet where the tweet_id is a unique integer identifying the tweet, sentiment is either 1<br>(positive) or 0 (negative), and tweet is the tweet enclosed in "". Similarly, the test dataset is a CSV file of<br>type tweet_id, tweet. Please note that CSV headers are not expected and should be removed from the<br>training and test datasets. used Anaconda distribution of Python for datasets for library requirements<br>specific to some methods such as keras with TensorFlow backend for Logistic Regression, MLP, RNN<br>(LSTM), and CNN. and xgboost for XGBoost. Usage of preprocessing, baseline, Naive Bayes, Maximum<br>entropy, Decision Tree, random forest, multi-layer perception etc are implemented.