Pinned

1. time-gebm time-gebm Public

   Two Generalized Energy Based Models for time-series generation. Thesis for MSc Computational Statistics and Machine Learning program at UCL.

   Python 3

2. rubix rubix Public

   Deep Reinforcement Learning based Rubik’s Cube solver written in JAX, Haiku and RLax. Developed a custom environment supporting cubes of variable size. Currently developing support for DQN, QR-DQN …

   Python 1

3. port-rl port-rl Public

   Reinforcement Learning based Portfolio Manager. Here the agent allocate portions of a fund to different financial products with the goal of maximising returns. Currently supports DDPG and PPO.

   Python 3

4. Multi-Task-Genre-Classification Multi-Task-Genre-Classification Public

   Multi-task learning models built for song genre classification made as a group project for the Natural Language Processing course at UCL. Here we consider both BERT and CNN based architectures.

   Python

5. Multi-Task-Learning-for-Image-Seg Multi-Task-Learning-for-Image-Seg Public

   Forked from yakovsushenok/multi-task-learning

   A multi-task learning model built for image segmentation made as a group project for the Deep Learning course at UCL

   Python