Risk Predictor

"Prediction of Risks Associated with Mass Corona Vaccination"

PROBLEM

There is a trend among people not to get COVID-19 vaccinations.Society is doubtful of getting COVID-19 vaccines due to spreading opinions, various myths, and fear of getting side effects. However, there is no research being done into the vaccination's side effects or the causes of illnesses and deaths.



This project will look into the specific criteria or risks that come with vaccinations.

OUR OBJECTIVE

To derive and validate risk prediction algorithms to estimate the risk of covid-19 related to side effects after the vaccination of people By, Creating a comprehensive data set with data mining and other techniques Investigating the effects of Demographic Factors, Health Conditions, Genetical influences, and Habitual influences for risks associated with mass corona vaccination. Developing an Interactive Web site with all the data, statistics, analytics, and visualization. So, this will help people to get an idea to check whether to take the vaccines or not.



SOLUTION ARCHITECTURE



To predict the risks associated with mass corona vaccination we are going to analyze different kinds of side effects (Fever,Itching,Coughing,Joint pain,Headache,Muscle pain,Swelling , Redness etc) with the following parameters,

  • Age
  • Gender
  • Height and Weight
  • Vaccine Type
  • Blood Group
  • Living Area
During the above analysis process first, we will try out different machine learning models, and then we will choose the best model by comparing the accuracy of each model. Our machine learning model will do the predictions on above mentioned side effects and finaly give a probability of being affected with a particular side effect. Finally, a web application that is combined with the best machine learning model will be developed. A database is maintaied at the backend at it will store the predictions along with the details provided by the user. Those data will be then used as another dataset.


Use case diagram



In our system (web application) there are two main actors. The person who visits our website (CUSTOMER) is the primary actor. The machine learning model is the secondary user of our system. There are four main use cases in our system and they are shown below

  1. submit a form : The user has to submit a form and that form will take some details of the user. Once the user clicks on the submit button the web application will check the validity of the entered data, so it is an 'include' relationship. If the details are incorrect the web application will display an error message, so that is an 'extend' relationship. Finally the data entered by the user will be fed into a ML model to do predictions.
  2. view graphs on analyzed data : Once the user clicks on the submit button he or she will be navigated to another page and it will show the graphs on analyzed data
  3. do predictions : ML model will predict the percentage of having a particular side effect for a person, according to the prediction algorithm, past data set and the details entered by the user.
  4. view predictions : The user will be able to see the percentage of being affected with a particular side effect.



ER diagram


SOFTWARE DESIGN

Our web application consists of 6 main pages.Those pages are shown bellow with their functionality.

  • Welcome Page : This is the very first interface that the user can see.
  • About page : By going to this page the user will be able to get a clear idea about our web application.
  • Form page : when the user clicks on the 'click here to continue' button he or she will be navigated to this page and it will get some essential details from the user.
  • Overall analysis page :when the user clicks on the 'view Overall analysis' button he or she will be navigated to this page and it will show the graphical representation of analyzed data based on the already available data.
  • Dashboard : once the user clicks on the 'submit' button of the 'Form' page,he or she will be navigated to this page and it will show the prediction that are done by the machine learning model.
  • Contact Us: This page consists of the contact details of the members of our team.

UI Designs


Welcome Page


Content of the 'About Page'


Form page


Contact Us Page


Front End Validation

All the user inputs that are taken from the form, are checked before sending them to the server.


Valid inputs


Invalid inputs



Web App Demostration (To be updated .....)




Technologies used

MACHINE LEARNING DESIGN

This study aims to predict the occuring of side effects after the COVID-19 vaccine.

Work Flow


In this study, Four supervised machine learning models will be used to train the dataset and predict the output by using testing dataset:

Logistic Regression

Random forest

Support vector machine

K-nearest neighbor


Progress Review...


DataSet

Data Preprocessing



Train Test split

Linear Regression Model



TESTING


  • POSTMAN is a scalable API testing tool.
  • HTTP requests that are sent to the server in each api call, are tested using this tools.

  • Selenium is an open-source and a portable automated software testing tool for testing web applications
    		•

    API testing with POSTMAN


    TIMELINE

    TEAM

    Our Team Members

    Madush Dilshan
    E/17/040

    Shashini Upekha
    E/17/356

    Hasara Wijesooriya
    E/17/407

    Our Advisors

    Dr.Suneth Namal

    Dr.Upul Jayasinghe

    Nuwan Jaliyagoda

    Contact Us


    University of Peradeniya.

    Phone: +94 81 239 33 00
    Email: vc@pdn.ac.lk
    Web-site: http://www.pdn.ac.lk/

    Faculty of Engineering.

    Phone: +94 81 239 33 02
    Web-site: http://eng.pdn.ac.lk/

    Computer Engineering Department.

    Phone: +94 81 239 39 14
    Web-site: http://www.ce.pdn.ac.lk/