MAN KI BAAT !!
It is a depression and suicide prediction and detection model using machine learning ,and visualized in a web page developed in R ,html,css,java script building logistic regression and random forest.
Created on 1st February 2020
•
MAN KI BAAT !!
It is a depression and suicide prediction and detection model using machine learning ,and visualized in a web page developed in R ,html,css,java script building logistic regression and random forest.
The problem MAN KI BAAT !! solves
People can use it for detection weather they are having depression or not.Suppose someone is feelling that he /she is low by mind for somedays which actually the early stage of depression.nearly 4.4 million people is suffering fro m depresiion but they are still undetected!as no one is aware of it ,doctors generally takes a lot of time for the detection of this disease but weare solving this problem by this model where we are created smart doctor model where it wil ask fist patients some questions based on psychoanalysis,on the basis of which it will predict the person is having depression or not or may he/she have suicide or not ?mean while the site is designed in such a way that the person will get coonections to the doctors,lit up mind!!and get a thorough knowledge about the depression.
Hence the detection can be done self,increased awareness,and at early stage by simple questions!!
Challenges we ran into
Firstly when we were trainning the model we were facing problems to set the detection parameteres!!we fixed it by doing some research over it.finally error optimisation and validation of the model faced some issues hence we runned diff test cases to prevent over fitting of the model!!
Next in the frontend we faced problems to design the screen as we wanted to create the simpliest way to predict it without any monotony so we researched upon the existing models and produced ours in a beter way !!
next while in the front end we were facing boolean expression error while designing our buttons we solved it by utilizing node.js
Html question interface was facing problem to be called with hence it is solbved by linking it with linking it in webpage!!
Technologies used