Deep clustering with concrete k-means
This addresses the challenge of deep clustering for researchers and practitioners by enabling more efficient and integrated training without alternating optimization, though it appears incremental as it builds on existing deep clustering methods.
The paper tackles the problem of jointly learning k-means clustering and deep feature representations from unlabeled data, achieving this by developing a gradient-estimator using the Gumbel-Softmax reparameterization trick to optimize the canonical k-means objective end-to-end.
We address the problem of simultaneously learning a k-means clustering and deep feature representation from unlabelled data, which is of interest due to the potential of deep k-means to outperform traditional two-step feature extraction and shallow-clustering strategies. We achieve this by developing a gradient-estimator for the non-differentiable k-means objective via the Gumbel-Softmax reparameterisation trick. In contrast to previous attempts at deep clustering, our concrete k-means model can be optimised with respect to the canonical k-means objective and is easily trained end-to-end without resorting to alternating optimisation. We demonstrate the efficacy of our method on standard clustering benchmarks.