Deep Kernel Learning for Clustering
This addresses clustering problems for data scientists by providing a faster, more expressive alternative to spectral methods, though it appears incremental as it builds on existing deep clustering and kernel techniques.
The authors tackled the problem of discovering kernels for clustering sample data by proposing a deep learning approach that produces embeddings at least as expressive as spectral clustering. Their method outperformed state-of-the-art deep clustering and traditional approaches like k-means and spectral clustering across various real-life and synthetic datasets.
We propose a deep learning approach for discovering kernels tailored to identifying clusters over sample data. Our neural network produces sample embeddings that are motivated by--and are at least as expressive as--spectral clustering. Our training objective, based on the Hilbert Schmidt Information Criterion, can be optimized via gradient adaptations on the Stiefel manifold, leading to significant acceleration over spectral methods relying on eigendecompositions. Finally, our trained embedding can be directly applied to out-of-sample data. We show experimentally that our approach outperforms several state-of-the-art deep clustering methods, as well as traditional approaches such as $k$-means and spectral clustering over a broad array of real-life and synthetic datasets.