Deep Online Probability Aggregation Clustering
This work addresses instability and computational issues in deep clustering for machine learning practitioners, offering a novel method that improves performance, though it appears incremental as it builds on existing deep clustering approaches.
The paper tackles the instability and computational burden in deep clustering by proposing a centerless clustering algorithm called Probability Aggregation Clustering (PAC), which formulates clustering as an optimization problem to align probability and distribution spaces, and integrates it into a deep framework (DPAC) that outperforms state-of-the-art methods in experiments.
Combining machine clustering with deep models has shown remarkable superiority in deep clustering. It modifies the data processing pipeline into two alternating phases: feature clustering and model training. However, such alternating schedule may lead to instability and computational burden issues. We propose a centerless clustering algorithm called Probability Aggregation Clustering (PAC) to proactively adapt deep learning technologies, enabling easy deployment in online deep clustering. PAC circumvents the cluster center and aligns the probability space and distribution space by formulating clustering as an optimization problem with a novel objective function. Based on the computation mechanism of the PAC, we propose a general online probability aggregation module to perform stable and flexible feature clustering over mini-batch data and further construct a deep visual clustering framework deep PAC (DPAC). Extensive experiments demonstrate that PAC has superior clustering robustness and performance and DPAC remarkably outperforms the state-of-the-art deep clustering methods.