The problem Me&Me solves

Our team has chosen this problem statement because, we ourselves being students, are
very well aware of the state of mental health of our colueges and also how this pandemic has
added to worsening it further.
Therefore, we have come up with the application that helps students combat their ill mental health in a quicker and much easier way.

Challenges we ran into

Android Application problems:

1. For the app development part making the post request with around 80 values was a troublesome task , used map data structure to handle it .
2. Proper heirarchy of firebase realtime database was required for effieciently adding and updating the data.
   ML Problem :
3. Data set was vast, it was having 220 features so, after linear regression, it was overfitting data, so, we had to try regularization
4. In regularization, some features were removed manually on the basis of correlation value and some statistical value.
5. We had depression as 0 and 1 only i.e classification, but we want to have a relative percentage for guardian to relate the behavior of student so, using correlation with this 3 values as an multiplier, we did this, and we prepared final depression score