A collection of random thoughts and experiments on machine learning topics.

Super resolution

Super Resolution

A qualitative and quantitative comparision of three various architectures that can be used as generators for super-resolution.

Read More

beating mnist

Beating MNIST

An iterative notebook that highlights how by applying recent best practices i nimage processing you can achieve state-of-the art results on the MNIST dataset.

Read More


Some introductory material on the scikit-learn framework.

Read More