|
Activity Number:
|
198
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 7, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistical Graphics
|
| Abstract - #306754 |
|
Title:
|
Of Forests, Trees, and Logs: Application and Visualization of Tree-Based Methods for Large Data
|
|
Author(s):
|
Simon Urbanek*+
|
|
Companies:
|
AT&T Labs-Research
|
|
Address:
|
Shannon Laboratory, Room D281, Florham Park, NJ, 07932,
|
|
Keywords:
|
large data ; tree-based models ; classification and regression trees
|
|
Abstract:
|
Tree-based methods became mainly popular due to their versatility and interpretability. However, some tree-based approaches are computationally intensive, which raises the question of their applicability to large data. Moreover, visualization and interpretability of tree models can be impaired by model complexity. In this paper we discuss the performance of tree models on large data, their inherent properties in such an application and present visualization techniques that help to enhance interpretability and extract information from both complex tree models and tree models for large data.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2006 program |
