Analyzing over 40000 movie reviews and training a neural network to successfully classify them
Creating a Sentiment Analysis Classifier using Python & PyTorch.
Goals:
Accurately predict the sentiment of a review Use a RNN Neural Network to analyze and classify the text. Use hugging face to upload a good working classifier that can be used by anyone. Aim to get an accuracy of atleast 85% to ensure reliability in results. First Completion Notes:
Learned about RNNs (GRU mainly and a bit about LSTM) Learned how to manipulate large datasets (40000 reviews) and tokenize them Successfully built a custom model that brings a reliable accuracy of 88.2% New Goals:
Make a usable application using huggingface (in progress)