Overview / Usage

Scratch4ML is a toolkit for making programming more accessible and fun. With MIT's Scratch as the inspiration behind this project we came up with a visual representation of Machine Learning blocks. Having a visual language means you don't have to focus on learning a syntax to start programming. It's much more easier to get into a new technology if it's easy to understand and begin with. Blocks as an abstraction can more easily capture concepts of programming. Scratch4ML contains widely used ML algorithms as blocks. It also comes with dataset like MNIST and California Housing.

Methodology / Approach

1. The user selects the dataset to be used.
2. The user drags and drops in the different components of the machine learning API. It can be a standalone ML algorithm or an operation.
3. User clicks on Run.
4. This runs the algorithm on the selected data and output stats.

Technologies Used

1. Python
2. JavaScript
3. scikit-learn

Repository

https://github.com/xprilion/novato

Other links

• Scratch