EPITECT is an initiative taken by us to forecast/predict the number of patients infected with Novel Corona Virus COVID-19. Our Model is a Machine learning based model that predicts the number of patients infected with corona, the recovery rate and the number of deaths that will be caused due to corona in the upcoming days in the future.
The new strain of Coronavirus has had a worldwide effect. It has affected people from different countries. The dataset provides, a time series data tracking the number of people effected by the virus, how many deaths has the virus caused and the number of reported people who have recovered.
Data is coming from https://github.com/CSSEGISandData/COVID-19 updated daily. We have normalized data a bit - unpivoted and transferred dates to be more machine readable. Alongwith that we’ve used dataflows to process and normalize the data.
The training is done through Artificial Neural Networks. The dates, longitude and latitude are label encoded and one hot encoded and are fed into a artificial neural network having two hidden layers and an output layer having the linear activation function. The network is then used to predict the number of cases in the future by changing the date data field. With the help of Drop-down menu you select the locations whose prediction you want and accordingly the results are displayed (presently due to lack of data there’s a limitation to this attribute, but in future with sufficient available data it can be proceed easily).
Basically the UI displays two types of result :-
• The World-wide prediction of the disease, with the help of global map.
• Second is the region-specific prediction based on the region selected from the drop-down It shows various comparisons among different predicted rates in form of Line-Charts as well as Bar-Charts. These comparisons can be studied for further analysis by the concerned authorities.
We assume you have python 3.0+ installed.
Create a python virtual environment:
• pip install virtualenv
Then create a directory where you can keep all your virtual environments.(like venvs) then cd to that directory
• virtualenv epitect : this will be your virtual enviroment for all related packages.
• git clone https://github.com/sgbro/HealthCare-EPITECT.git
• cd to the cloned repository
• install the required libraries.
NOTE: activate your virtual environment before this so that it gets installed in that environment.
• run the app.py file so that the project will run in your browser locally.
1. The accuracy of model is limited by some attributes such as life styles, population dynamics etc. which can be implemented in future with proper availability of the datasets.
2. Certain additional data has to be requested to govt. as well as healthcare institutions that is not currrently available in the public domain.
3. The unstructured data or raw data especially in the field of healthcare makes it difficult to use more and more structured data.
4. Various other factors such as population density of an area, Economic profile (GDP, GDP PPP), Two weeks' data, Reported cases, Vector agent occurences, proximity with other populated regions greatly affects the prediction results which can be integrated with the model in future but not now due to lack of datasets.
