Oracle-Preserving Latent Flows
This work addresses the challenge of identifying symmetries in data for machine learning applications, but it appears incremental as it builds on existing symmetry discovery methods with specific extensions.
The paper tackled the problem of discovering multiple continuous symmetries in labeled datasets by developing a deep learning method using neural networks and a specialized loss function, achieving demonstration on the MNIST dataset.
We develop a deep learning methodology for the simultaneous discovery of multiple nontrivial continuous symmetries across an entire labelled dataset. The symmetry transformations and the corresponding generators are modeled with fully connected neural networks trained with a specially constructed loss function ensuring the desired symmetry properties. The two new elements in this work are the use of a reduced-dimensionality latent space and the generalization to transformations invariant with respect to high-dimensional oracles. The method is demonstrated with several examples on the MNIST digit dataset.