Why do modern deep neural networks (DNNs) perform so well on previously unseen test data, even when their number of weights is much larger than the number of data points in the training set? This question keeps puzzling many theorists and practicioners doing Deep Learning (DL), in particular those who …
Machine learning can detect classical topological defects in materials
Tensor networks are a powerful tool for compressing quantum wavefunctions developed in the physics community. Parameterizing a special class of models with tensor networks brings the full power of tensor networks to machine learning tasks.
How do quantum mechanics and machine learning fit together?