HaiderSultanArc / LinuxDistro4U

LinuxDistro4U is a Web Application that recommends you a Linux Distro based on your Responses on a set of Questions, using Machine Learning

LinuxDistro4U

Find the best Linux Distro for yourself

Desciption:

Choosing a Linux Distro can be difficult as there is a ton of options to go with. So, I made this little website. It's a Questionnaire based website that uses Machine Learning to predict the Linux Distro that will best suit you, based on your response. It makes the choice for you and gives a single best name of the distro for your priorities.

Visit Website here: LinuxDistro4U

Techologies used:

The project was divided into 2 parts:

1. Machine Learning Model

• For Machine Learning, I worked in Python Tensorflow . For my model to work on Browser, I used TensoflowJS that imported TensorflowLite Model from Google Storage.

2. Frontend Development

• For Frontend Development, I worked in ReactJS

3. Hosting and Data Storage

• I used Firebase for hosting and firebase realtime datbase for data storage

Things I learned:

• Collecting Data for Machine Learning
• Data analysis and Processing for Machine Learning
• Developing Machine Learning Models
• Deploying ML models
• Developing Frontend in ReactJS for ML Applications
• Dockerization
• Gitub Action CI/CD
• Using Google Cloud Console

Milestones:

Version 1.0:

• Create a Linux Distro Library with all their features (Only Distros that can be used as Desktop, are active for more than 10 years are included)
• Collect Data for Machine Learning
• Develop Machine Learning Model
• Develop pages for Website
• Develop Arc'Ion and his Speech System
• Develop Questionnaire System
• Develop Input Vector From responses
• Integrating Tensorflow.Keras Model into Website using TensorflowJS, host model on firebase
• Host version 1.0 on Firebase

Version 1.5:

• Upgrade to React 18
• Updrade to mui
• Move Linux Distro Library and Questions to Firebase
• Move TensorFlow Model to firebase using firebase-admin

Version 2.0:

• Setup CI/CD using Github Actions
• Develop a Python Backend for the Application
• Use TensorFlow Model with Python Backend
• Dockerize the solution
• Host version 2.0 on Vercel

Version 3.0:

• Include Linux Distros that are Active for 5 or more years and can be used as Desktop
• Re-Collect the Training Data
• Re-Train the Model and Deploy it
• Host version 3.0 on Vercel

Version 4.0

• Include Linux Distros that are Active for 5 or more years and are special purpose
• Re-Collect the Training Data
• Re-Train the Model
• Host version 4.0 on Vercel

Connect with me: