Sentiment-Analysis-Using-LSTMs

Dataset Used:

IMDB movie reviews by Keras. It consists of 25,000 labeled movie reviews on IMDB. Reviews sentiments are labeled as positive/negative. You can read more about it here -> https://keras.io/datasets/#imdb-movie-reviews-sentiment-classification

To run the project:

Just open the link to the colab notebook and you should be good to go. :)

Architecture:

Currently trying to improve the accuracy of the model, will update the final architecture here, once done.

Code Flow:

Will be updated once the project is finished.

Results:

Current model will the following stats
precision recall f1-score support

Negative       0.85      0.92      0.88     12500
Positive       0.91      0.84      0.87     12500

accuracy                           0.88     25000

macro avg 0.88 0.88 0.88 25000 weighted avg 0.88 0.88 0.88 25000

Accuracy: 0.87616

Recommended Settings:

• Go to the runtime tab drop down menu, click on Change runtime type
• select GPU to reduce the training time.