Welcome !

Hi there, I'm tulasi ram - Data Scientist [codeLOVEr]

• Know more about me ** Portfolio ** 👋

I'm a Data Science, Machine Learning, NLP, Deep Learning, Artificial Intelligence Enthusiast!!

• 🔭 I am a recent Graduate : [Want to Become A Data Scientist!]
• 🌱 I’m currently learning everything 🤣
• 👯 I’m looking to collaborate with other developers
• 🥅 2020 Goals: Improve and gain Knowledge on ML techniques
• ⚡ Fun fact: I love to travel, play video games, reading and writing articles

Connect with me:

• Let's stay connected linkedin
• Read my articles Medium
• For Introducing Skype
• Let's get Connect Instagram

Bussiness Objective

Twitter has become a large platform to extract data and can be used to solve different kinds of bussiness objectives.

• Customer behaviour analysis
• sentiment analysis
• AI chatbots
• Recommendation system, etc

In our case, we collect different kinds of tweets with these keywords Depressed, Depression, Hopeless, Lonely, Suicide, Antidepressant Antidepressants from twitter and analyse to depression prediction and it appears that this solution is significant enough to have solved the difficulty.

Data Collection:

Tweets collected on Linux system commands using Twint tool. This tool is a magical for developers to collect data for thier desired use cases.

• Random tweets that do not necessarily indicate depression and tweets that demonstrate that the user may have depression and/or depressive symptoms.
• A dataset of random tweets can be sourced from the Sentiment140 dataset available on Kaggle

https://drive.google.com/drive/folders/1z-PrTTT6u3xciSUc0eZQRfQa4qn09urc?usp=sharing

Data Exploration & Data visualisation

• Words Frequency
• Characters Frequency
• Most common words
• word cloud

Model Evaluation and Validation

Hence it is a binary classification model, Accuracy and loss are recorded and visualized and compared to a benchmark logistic regression model.

conclusion

The final model proves to be far more accurate than the benchmark model. The benchmark model, run on the same data for the same number of epochs, shows an accuracy of approximately 64%, while the final model has an accuracy of approximately 97%. This proves to be a much more robust and effective model for depression prediction and it appears that this solution is significant enough to have solved the difficulty of effectively analyzing Tweets for depression.

For more information get into my article on medium

https://medium.com/swlh/detecting-depression-in-social-media-via-twitter-usage-2d8f3df9b313