Bayesian rose trees
Hierarchical structure is ubiquitous in data across many domains. There are many hierarchical clustering methods, frequently used by domain experts, which strive to discover this structure. However, most of these methods limit discoverable hierarchies to those with binary branching structure. This l...
| Päätekijät: | , , |
|---|---|
| Aineistotyyppi: | Journal article |
| Julkaistu: |
2010
|
| _version_ | 1826285977171132416 |
|---|---|
| author | Blundell, C Teh, Y Heller, K |
| author_facet | Blundell, C Teh, Y Heller, K |
| author_sort | Blundell, C |
| collection | OXFORD |
| description | Hierarchical structure is ubiquitous in data across many domains. There are many hierarchical clustering methods, frequently used by domain experts, which strive to discover this structure. However, most of these methods limit discoverable hierarchies to those with binary branching structure. This limitation, while computationally convenient, is often undesirable. In this paper we explore a Bayesian hierarchical clustering algorithm that can produce trees with arbitrary branching structure at each node, known as rose trees. We interpret these trees as mixtures over partitions of a data set, and use a computationally efficient, greedy agglomerative algorithm to find the rose trees which have high marginal likelihood given the data. Lastly, we perform experiments which demonstrate that rose trees are better models of data than the typical binary trees returned by other hierarchical clustering algorithms. |
| first_indexed | 2024-03-07T01:36:56Z |
| format | Journal article |
| id | oxford-uuid:958397f6-58c5-400e-b730-aad8152a1b56 |
| institution | University of Oxford |
| last_indexed | 2024-03-07T01:36:56Z |
| publishDate | 2010 |
| record_format | dspace |
| spelling | oxford-uuid:958397f6-58c5-400e-b730-aad8152a1b562022-03-26T23:46:41ZBayesian rose treesJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:958397f6-58c5-400e-b730-aad8152a1b56Symplectic Elements at Oxford2010Blundell, CTeh, YHeller, KHierarchical structure is ubiquitous in data across many domains. There are many hierarchical clustering methods, frequently used by domain experts, which strive to discover this structure. However, most of these methods limit discoverable hierarchies to those with binary branching structure. This limitation, while computationally convenient, is often undesirable. In this paper we explore a Bayesian hierarchical clustering algorithm that can produce trees with arbitrary branching structure at each node, known as rose trees. We interpret these trees as mixtures over partitions of a data set, and use a computationally efficient, greedy agglomerative algorithm to find the rose trees which have high marginal likelihood given the data. Lastly, we perform experiments which demonstrate that rose trees are better models of data than the typical binary trees returned by other hierarchical clustering algorithms. |
| spellingShingle | Blundell, C Teh, Y Heller, K Bayesian rose trees |
| title | Bayesian rose trees |
| title_full | Bayesian rose trees |
| title_fullStr | Bayesian rose trees |
| title_full_unstemmed | Bayesian rose trees |
| title_short | Bayesian rose trees |
| title_sort | bayesian rose trees |
| work_keys_str_mv | AT blundellc bayesianrosetrees AT tehy bayesianrosetrees AT hellerk bayesianrosetrees |