GitHub

Implement MLP with only numpy module, without deep learning tools (e.g. PyTorch, Tensorflow, ...)

Simply run python ./train.py to try our numpy MLP. We perform the MNIST handwritten numbers classification task to conduct the experiment. You can also use arguments to adjust the training process: ...
GitHub

AkiRusProd/numpy-nn-model

A custom CPU/GPU, torch-style machine learning framework with tape-based automatic differentiation. Built on NumPy for CPU and CuPy for GPU support. Layers - Neural network layers including ...
IEEE

A Review of SiC Power Module Packaging Technologies: Challenges, Advances, and Emerging Issues

Abstract: Power module packaging technologies have been experiencing extensive changes as the novel silicon carbide (SiC) power devices with superior performance become commercially available. This ...
UCL

Module selection

Welcome to UCL! If you are a new student starting on a new programme of study for January 2026, see here for information about module selection.