This repository contains an interactive web-based platform designed to visualize machine learning algorithms step-by-step.
The main goal of this project is to help students, beginners, and educators understand the internal workings of common ML algorithms through clear, animated, and intuitive visual simulations.
The website is hosted on GitHub Pages:
🔗 https://avir.github.io/ml/
During my machine learning course, I realized that many algorithms are easy to understand theoretically but much harder to visualize.
For example, in algorithms like K-Means Clustering, concepts such as centroid initialization, distance calculation, and iterative updates become far clearer when seen visually.
So I decided to build this platform to:
- Provide interactive demonstrations of ML algorithms
- Make learning ML more intuitive and engaging
- Help classmates and juniors learn through visualization
- Enable faculty members to use these simulations in class
- Build a long-term project that expands as I learn new algorithms
Each algorithm will have:
- A dedicated page
- Interactive controls (inputs, sliders, dataset options)
- Step-by-step visualization of the algorithm’s internal process
- Dynamic plots and animations
- Clear explanations of each step
The first implemented algorithm is:
- Random centroid initialization
- Euclidean distance calculation
- Cluster assignment
- Centroid recomputation
- Iterative visualization until convergence
As I progress through my ML course, I will continue adding more algorithms, such as:
- K-Nearest Neighbors (KNN)
- Linear Regression (Gradient Descent visualization)
- Logistic Regression
- PCA (Dimensionality Reduction)
- DBSCAN
- Decision Trees
- Support Vector Machines
- And more…
This repository will grow into a complete ML Algorithm Interactive Learning Hub.
The project is entirely client-side and hosted using GitHub Pages, making it easy for anyone to access and explore directly from their browser.
Website:
🔗 https://avir.github.io/ml/
Suggestions, improvements, or new feature ideas are always welcome.
Feel free to open an issue or submit a pull request.
This project is open-source and available under the MIT License.