A Consciousness-Inspired Planning Agent for Model-Based Reinforcement Learning
Posted on22/11/2021
Editorial Note: This blog post was originally published on the author’s personal website and has been reviewed and adapted for the Mila... Read More
This Climate Does Not Exist: Picturing impacts of the climate crisis with AI, one address at a time
Posted on13/10/2021
Editorial Note: This article was written with support from the team behind This Climate Does Not Exist. See end of post for... Read More
Flight-SEIR: Incorporating Flight Data to Improve Epidemiological Modelling and Disease Outbreak Prevention
Posted on03/08/2021
A modified version of the standard SEIR model that incorporates COVID-infected flights in and out of Canada could enable early detection of outbreaks, more accurately estimate the reproduction number of the disease and better evaluate the impact of travel restrictions and the implications of lifting these measures.
Introducing SpeechBrain: A general-purpose PyTorch speech processing toolkit
Posted on28/04/2021
What is SpeechBrain? SpeechBrain is an open-source and all-in-one speech toolkit. It is designed to make the research and development of neural... Read More
Improved Deep Learning Workflows Through Hyperparameter Optimization with Oríon
Posted on03/12/2020
A collaboration between Mila and IBM for the development of Oríon Hyperparameter optimization (HPO) procedures are crucial for learning algorithms to achieve... Read More
La-MAML: Look-ahead Meta-Learning for Continual Learning
Posted on19/11/2020
We propose La-MAML, a fast and online meta-learning algorithm for continual learning from streaming data, with inbuilt learning rate modulation.
Towards Precision Medicine: Understanding Inference and Prediction Divergence in Biomedicine
Posted on08/10/2020
In a future of single-patient prediction from big biomedical data, it may become central that modeling for inference and modeling for prediction are related but importantly different.
A collaboration with Stony Brook Medicine to Build a COVID-19 Severity Prediction Tool
Posted on20/07/2020
Mila collaborates with Stony Brook Medicine. A team led by Joseph Paul Cohen builds a public dataset and models to predict severity of COVID-19 pneumonia. Initial results are promising and the tools are ready for more evaluation.
Learning Better Representations by Interpolating Hidden States
Posted on02/07/2020
Manifold Mixup is a simple and easy-to-implement regularization method that improves generalization of deep learning models.