This project performs sentiment analysis on movie reviews. The first version uses a Multilayer Perceptron (MLP) model to classify reviews as positive or negative.

In the advanced version, a Bidirectional LSTM model replaces the MLP, aiming to improve classification accuracy and performance. It demonstrates how different model architectures impact sentiment analysis results.

In conclusion, both models produced very similar results. Interestingly, I found that many others on Kaggle achieved comparable outcomes, suggesting possible dataset bias or trends that are challenging for models to capture. Overall, these models demonstrate a moderate level of performance on this dataset.