Today , important of human health , we need solve two major problems. First one spreading Risk Factors of Covid-19 we must need to find and reduce and Second one Risk Factors of Covid-19 patients. Our inspiration simply we need to overcome pandemic era.
Last one month i had search valid data-set for Covid-19 patients. But the meanwhile i have understood most of the data-base used for tracing and counting. so We have to planned a common Machine Learning tool for Reduce the Corona Spreading level based on their foods, medical history, social distance parameters. The Risk Prediction patients based on their clinical check list. we have to plan a common tool for Prediction.
What it does
It's a common Tool GUI, The name itself , if you have local area patient checklist/attributes then easy to Train the Data using this GUI and finally predict the Risk level of Unknown Patients.
Train Data(Risk Level Known Patient) --> Feature Normalization --> optimized Feature Selection ---> Train The Classifier Test Data(Risk Level Unknown)--> Normalization---->selected Features---> Predict the Result with the help of Trained Classifier. In simple-ways, if you have your local area patient check list from that we have find the Risk prediction/Infection Chance. Most importantly , if we do in correct way, we will easy to find which are the parameters like foods and social distance type etc reduce the Covid-19 pandemic.
How I built it
using MATLAB GUI . Design and Evaluate the Performance
Challenges I ran into
Original Dataset Collection is complex. In this Data-set we have achieved 87% Training Accuracy and More than 75 % Testing Accuracy.
Accomplishments that I'm proud of
Initial Version 1.0 Task Completed
What I learned
We have already entered Covid-19 pandemic , The Process like Reverse Engineering by using Machine Learning reduce the impact of Covid-19.
What's next for Covid-19 Patient/Non-Patient Risk Prediction using AI
here up-to now , we have completed Simple level ML Tool for Risk-class Prediction. In Future we need to update both Risk Class and Values by using Fuzzy and so many AI Techniques.