MIT researchers uncover the structural properties and dynamics of deep classifiers, offering novel explanations for optimization, generalization, and approximation in deep networks.
The method enables a model to determine its confidence in a prediction, while using no additional data and far fewer computing resources than other methods.
A new experiential learning opportunity challenges undergraduates across the Greater Boston area to apply their AI skills to a range of industry projects.
MIT researchers developed and studied a customized AI training program for users with varied backgrounds, which could be delivered across large organizations.
Stefanie Jegelka seeks to understand how machine-learning models behave, to help researchers build more robust models for applications in biology, computer vision, optimization, and more.