Clustering with Bregman Divergences: an Asymptotic Analysis
Authors: Chaoyue Liu, Mikhail Belkin
NeurIPS 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we verify our results, especially centroid s location distribution Eq.(24), by using the Bregman hard clustering algorithm. Figure 1 shows, in the first row, the theoretical prediction of distribution of centroids, and in the second row, experimental histograms of centroid locations for different Bregman quantization problems. |
| Researcher Affiliation | Academia | Chaoyue Liu, Mikhail Belkin Department of Computer Science & Engineering The Ohio State University |
| Pseudocode | No | The paper does not contain pseudocode or a clearly labeled algorithm block. It references 'Algorithm 1 in [3]', but that is external to this paper. |
| Open Source Code | No | The paper does not provide any statement or link indicating the release of open-source code for the methodology described. |
| Open Datasets | No | The paper describes sampling data points from a uniform distribution ('Suppose the density P is uniform over [0, 1].', 'The density P = U([0, 1]2)') for its experiments. It does not refer to a publicly available dataset with specific access information (link, DOI, citation). |
| Dataset Splits | No | The paper does not provide specific training/validation/test dataset splits. It discusses sampling data and applying clustering, but no traditional data splitting is described. |
| Hardware Specification | No | The paper does not provide any specific hardware details such as CPU/GPU models, memory, or cloud computing resources used for experiments. |
| Software Dependencies | No | The paper mentions applying 'standard k-means', 'Kullback-Leibler clustering', and 'norm-like clustering' algorithms, but does not provide specific version numbers for any software or libraries used (e.g., Python version, library versions). |
| Experiment Setup | Yes | We set number of clusters k = 81, and apply different versions of Bregman hard clustering algorithm on this sample: standard k-means, Kullback-Leibler clustering and norm-like clustering. In addition, we only verify r = 1 cases here, since the Bregman clustering algorithm, which utilizes Lloyd s method, cannot address Bregman quantization problems with r = 1. |