|
| | | |
We wish you a happy and safe holiday season and all the best for 2025
Non-Invasive Attributes Significance in the Risk Evaluation of Heart Disease Using Decision Tree Analysis
Shouman, M. and Turner, T.
Non-invasive attributes are low cost, easy to identify
attributes that can have significant influence in the risk
evaluation of heart disease. This research formulates a
combination of non-invasive attributes that can be
used in the risk evaluation of heart disease. It uses the
decision tree data mining technique to identify the
significance of different single, combined, and
calculated non-invasive attribute combinations in the
risk evaluation of heart disease against the benchmark
Cleveland heart disease dataset and a larger Canberra
hospital heart disease dataset. For each non-invasive
attribute combination, 10-fold cross-validation is
applied to ensure reliable performance measures.
Different equations of non-invasive attributes on the
Canberra dataset show that the best combination in the
risk evaluation of heart disease is age, sex, resting
blood pressure and Rohrer’s Index equation with a
mean accuracy and standard deviation of 73.8% and
4.9% respectively. |
Cite as: Shouman, M. and Turner, T. (2015). Non-Invasive Attributes Significance in the Risk Evaluation of Heart Disease Using Decision Tree Analysis. In Proc. Thirteenth Australasian Data Mining Conference (AusDM 2015) Sydney, Australia. CRPIT, 168. Ong, K.L., Zhao, Y., Stone, M.G. and Islam, M.Z. Eds., ACS. 185-193 |
(from crpit.com)
(local if available)
|
|