Lung Cancer Prediction using Machine Learning

This project is a survey assessment form and admin dashboard built using React on Next.js, and python ml libraries. The main goal of the application is to allow users to find out their lung cancer risk level, and to showcase our database learning throughout this module. We are using a SQL databse hosted on PlanetScale (cloud provider), with Prisma object relational mapper to support our CRUD serverless function api calls. This project is built and managed by Team37 for CSC2008-Database module.

Team37 Members

Name	Student ID	Github Username
Dylan Tok	2101372	@dthx2710
Chong Wei Bing	1802754	@cwb07
Xu Xueli	2101812	@xx-ue
Teo Leng	2102311	@tlengz
Lim Wei Jie	2101242	@penguiny
Jared Teo	2101979	@JaredTeo311

Instructions to run the program locally

• Clone the repository

  git clone

• Install Dependencies & Packages

  npm install

• Initialize python venv for prediction model api. Use initlinux instead if on Linux OS

  npm run init or npm run initlinux

• Create a .env file in the working directory
  • You can use the .env.sample file as a template
  • Update OS_ENV environment key to LINUX if on Linux OS for ml scripts to run properly
• Build the program

  npm run build

• Run the program

  npm start Open http://localhost:3000 with your browser to see the result.

Usage

• Risk Assessment Survey
  • Fill in your details and factors, range of minimum number is good/healthy, maximum number is bad/unhealthy
  • Submit and view your results from our prediction model
  • Your results be added to the database anonymously
• Admin Dashboard
  • Admin account details are in Users table of the database, but here are some test accounts to login:

    Username	Password
    art	8NAADrwQ1nS5
    beverie	ozJxVgU6
    cazzie	o4wrmXLeRVM
    cynthea	QnskuXu
    dolicon	doliconnn

  • Admin dashboard provides basic data visualization and a table view with filters and edit/delete features

Screenshots

Home Page

Survey Results

Admin Dashboard

Data Table and Filters

Others

• Database visualization GUI Tool

  Run npx prisma studio

• Planetscale (Cloud SQL Database) performance analytics
  1. Go to Planetscale
  2. Login using test account credentials:
     • Email: [email protected]
     • Password: csc2008team372023
  3. Go to project dashboard
  4. Navigate to Insights tab
  5. View analytics (Query latencies, Qps, Rows r/w, etc)
• Kaggle Dataset Used

  https://www.kaggle.com/datasets/thedevastator/cancer-patients-and-air-pollution-a-new-link