Project 2

Classifying reviews from an IMBD dataset as positive or negative.

See requirements.txt for required packages.

Linear Models

• Unzip training and test data under data.
• Run src/train_models.py to train all models and obtain testing and validation accuracy. All models get saved under data/model/ after training.
• Use the template in src/test_model.py. Replace with the name of the model you want to test and enter a name for the CSV file.
• CSV file containing results will be stored under data/result/.

LSTM

• Unzip training and test data under data.
• Run src/lstm.py to train model.
• Model will be saved under data/model/lstm.h5.
• Use Keras' load_model function to load the trained model and generate your predictions.