Sensitivity analysis of the topology of classification trees

Download
Author
Kobayashi, Izumi.
Date
1999-12Advisor
Buttrey, Samuel E.
Metadata
Show full item recordAbstract
The use of classification trees is one of the most widely used techniques in classification. It is well known that classification trees are not stable in their topology, in contrast to their robustness with respect to misclassification rate. This thesis defines a measure that compares the topology of two trees and studies how a tree's topology changes when the dependent (y) variable or the independent (x) variables are perturbed. This allows us to examine the "robustness" of tree topology under perturbation and to compare it to the robustness with respect to the misclassification rate under the same perturbations. We show that the tree topology can change significantly even for small perturbations in many sets of data. This suggests that even small measurement errors in the variables can affect the tree topology greatly. Because data are often measured with error, it follows that splitting rules in trees may not be suitable for use in making policy decisions. We propose a measure for tree topology, and show that tree topology changes faster than the misclassification rate does under mild perturbations. This finding formalizes the concept that tree models are more stable in terms of misclassification rate than in terms of topology.
Collections
Related items
Showing items related by title, author, creator and subject.
-
Search for a malevolent needle in a benign haystack
Gaver, D.P.; Jacobs, P.A.; Glazebrook, K. D. (2009);A domain contains a number w of non-hostile White (W) individuals: humans, vehicles, ships. A hostile Red (R) individual enters the domain and travels through the domain towards targets. If R reaches an attractive valuable ... -
The Potential Misinterpretation Problem of Residual Based Models Due to Industry Misclassification
Wang, Chong; Shao, Lili; Zhang, Chuan (MAG Scholar Marketing in Asia Group, 2017);Accounting researchers frequently employ industry-specific residual based models to draw inferences. Examples include discretionary accruals in Jones (1991) and modified Jones model (1995), and accruals quality in Francis ... -
Color and Edge-Aware Adversarial Image Perturbations
Bassett, Robert; Graves, Mitchell; Reilly, Patrick (ArXiv, 2021-03-22);Adversarial perturbation of images, in which a source image is deliberately modified with the intent of causing a classifier to misclassify the image, provides important insight into the robustness of image classifiers. ...