The programming assignments cover topics that are important for performing research on combining machine learning and robots. They will cover:

• Behaviour cloning and imitation learning
• World Models and Planning
• Model-free (PPO) and model-based RL methods (DDPG or PETS).
• Exploration and pretraining methods, such as HRL and goal-conditioned RL.
• Offline reinforcement learning and exploration
• Scaling DeepRL algorithms

The assignments are also designed to familiarize students with the software needed to perform research:

• Deep learning libraries, such as PyTorch
• Hardware constraints when working with real robots (power, compute, mechanical limits)
• Distributed computing for running proper experiments
• Visualization and analysis (the most important part)

If we have five assignments, you can drop your lowest mark.

Late Options:

• You have 7 late days you can use for assignments
• Your grade on the assignment may be reduced by 10% per day.

— En français —

Les travaux pratiques de programmation couvrent des sujets importants pour mener des recherches sur la combinaison de l'apprentissage automatique et des robots. Ils aborderont :