Depression-Detection-using-a-Hybrid-CNN-LSTM-Layer

This project aims to develop a hybrid deep learning model to classify whether a given text is depressive or not.

DATASET USED

Sentiment 140 https://www.kaggle.com/datasets/kazanova/sentiment140

LAYERS USED

Embedding + conv + batchnormalization + maxpooling + lstm + maxpooling.

EMBEDDING FILE

Google embedding file https://www.kaggle.com/datasets/suraj520/googlenews-vectors-negative300bingz-gz-format.

PREREQUISITES

Download the embeddings and dataset into the root folder. Install the necessary packages from requirements.txt.

RESULTS

check results.png, network architecture.png, and outline.png for more details regarding the implementation of the model.

BONUS TIP

All the code is available in pro_final.ipynb file.