Kiyo Kunii
London, UK
Kiyo Kunii
- Symbolic Reinforcement Learning using Inductive Logic Programming
A new reinforcement learning algorithm that applies inductive logic programming to achieve data-efficient and explainable learning, using Python and Answer set programming.
Imperial College London
MSc in Computing Science
- Deep Learning for Product Recognition at Ocado
・Worked with Ocado machine learning engineers to develop a 3D rendering training generator and computer vision model using Convolutional Neural Network (CNN). ・The software can generate 3D training images of Ocado products to train the CNN model so that it can classify Ocado products from any angle. ・Was responsible for overall management of the project as well as development of evaluation of the CNN model in Tensorboard. ・Awarded for "Corporate Partnership Programme Commendation for Group Project".