Apprendre en environnement contraint

(à venir)

Lectures

• ProtoNN: Compressed and Accurate kNN for Resource-scarce Devices

• Near-optimal sample compression for nearest neighbors

• Joint learning and pruning of decision forests

• Optimally Pruning Decision Tree Ensembles With Feature Cost

• Cost-complexity pruning of random forests

• A Compression Approach to Support Vector Model Selection

• Resource-efficient Machine Learning in 2 KB RAM for the Internet of Things

Modules

• EdgeML : entraîner un modèle lorsqu’on a peu de mémoire.