In today’s world stress has become a more
familiar word because of its disastrous impact on the
huge number of people worldwide. It is very important
to keep stress under control every time, as it is the
primitive reason for much major health issues. Some
people meditate to g e t r i d o f i t and others choose to
use medicines to control their stress levels. Students also
found with very much stressed out because of
academics, projects, exams, and whatnot. There are
many ways through which one can check whether you
have stress or not. According to this situation, the
medical diagnosis system based on human physiology
becomes more requisite as compared to others. Human
physiology-based study plays a important character in
the detection of mental stress in persons. There have
also been eventual researches which are done on the
detection of stress based on facial emotions. To find out
whether stressed or not we need to see a doctor and get
checked, but it seems to be not practical at all times to
do so. In fact, in the era of digitalism, where everyone
has a smartphone there is a dearth of finding novel
ways through which we can make use of technology to
detect your stress levels automatically. There are
wearable devices that detect stress levels based on your
body activity. Many approaches aim for the detection of
stress through the use of wearable devices. The
approach that we are presenting in this project is
predicting stress through medical data of the patients
using random forest regression. Additionally, an
examination between oneself fabricated convolution
neural model and a portion of the pre-trained models
has been finished. This is another methodology and we
are getting very promising precision by utilizing
sufficient research experiments on 2000 irregular trees
in the model. The results achieved are the outcomes of
effectively anticipated with the accuracy utilizing the
model. The outcomes of this research can be useful in
directing the future which explor