Stars
Neural Networks tend to overfit. One way to overcome this problem is by considering a sequence of optimization problems, where we modify the loss function so that we minimize the distance to the pr…
Build MLP from scratch only using Numpy. Able to perform simple supervised learning problems in any dimension and with variable layer configurations. It can also be used for Auto-Encoder problems.
Automatically Creates a wordcloud from the artists appearing in a given Spotify-Playlist
Implementation of two phase field approaches for the surface reconstruction problem and shape space learning. One based of the Modica-Mortola theorem and the other based on Ambrosio-Tortorelli
Randomly generate two dimensional rounded polygonal shapes in HTML SVG