Large-Scale Clustering through Functional Embedding
Published on Oct 10, 20083804 Views
We present a new framework for large-scale data clustering. The main idea is to modify functional dimensionality reduction techniques to directly optimize over discrete labels using stochastic gradien