Predictex

Inspiration

According to the Anxiety and Depression Association of America, anxiety disorders are the most common mental illness in the U.S. They affect 40 million adults, or 18.1% of the population every year. Even though this disorder is treatable, only 36.9% of those suffering receive treatment, most likely due to a perceived societal stigma, with men far less likely to seek treatment.Quite often even those surrounding us are not aware about our mental health so a tool that can predict mental illness would be extremely beneficial.

What it does

This application asks users to answer a few question and by considering the responses given by them it displays the percentage chance of having mental illness.

How we built it

We trained 3 classification models to find the chances that a person is suffering from mental health issues namely using Logistic Regression, SVM and Xgboost. The dataset used was OSMI mental health in tech survey (Open Sourcing Mental Illness). And found the best accuracy was obtained using Xgboost (81.4%) .The models were deployed using django to perform predictions on the entered data.

Challenges we ran into

We could not find any standard dataset so we tried performing analysis on the survey based data.

Accomplishments that we're proud of

We could successfully deploy our application to heroku so any user can get to know about their chances of being ill.This application can be incorporated into software company websites where the employees can keep a check on their mental health.

What we learned

Through this project we have learnt to create a model in different methods and with the help of metrics we found the more accurate method.We learnt about analysing data to gain useful insights.