Mila > Team > Thomas Jiralerspong

Thomas Jiralerspong

Deep Learning
Generative Models
Natural Language Processing (NLP)
Recurrent Neural Networks
Reinforcement Learning (RL)
Music Generation
Student B.Sc., Student, McGill University

I am an undergraduate student in Honours Mathematics and Computer Science working with Professors Doina Precup and Blake Richards on adding planning to hierarchical reinforcement learning agents and on training reinforcement learning agents in environments with sparse rewards,

I previously trained a model to optimize mechanical ventilation using Conservative Q Learning.

My research interests include deep reinforcement learning, generative models, natural language processing and brain inspired artificial intelligence.