/Depression_detection_usin_CNNandLSTM

Primary LanguagePython

Depression_detection_usin_CNNandLSTM

In this project the tweets are extracted manually using tweets.py and collected over 20k+ tweets from twitter using different keyword into one CSV file (fina_output.csv). Tweets till 5000 are labelled manually and used as a dataset for this project.

Procedure to run through the whole project:

  1. Run tweets.py to collect tweet from the twitter.
  2. label them as depressed or not (can also use the database i used)
  3. Run the main.py file to run the model

model.py contains the architecture of the model used

reports are also attached as which algorithm worked with how much accuracy

Decision tree worked better comapring to other machine learning algo, with the accuracy of nearly 90%

CNN + LSTM outperformed and had an average of 96% accuracy