International Journal of Statistics and Reliability Engineering

No Publication Cost

Vol 10, No 2:

subscription

AUC Estimation and ROC Model Comparison in the Perspective of Generalized Exponential Distribution

Abstract

Binormal ROC model is one of the most widely used models due to the assumption that the random samples/observations/scores of two populations follow normal distributions. In the ROC literature, over the years, many researchers have worked on several bi-distributional ROC models. All these bi-distributional ROC models were developed based on the data pattern and the distributional fit, a few to mention as bi-gamma, bi-exponential and bi-half normal. In most of the practical situations, existing models may or may not be fit for any considered data. Hence, there will be in need of a new ROC model which can fulfill the distributional fit and account for the tail pattern of the data. With this, we made an attempt to propose a new form of ROC curve that underpins generalized exponential distribution with scale and shape parameters and later the same is compared with exponential based distributions. The proposed ROC model is supported with both simulated data sets and a real data set, namely APACHE IV. It is shown that the proposed ROC model accommodates the tail end behavior that in turn provided better accuracy than the other models under comparison.

Full Text

PDF

References

1. D. Bamber; The Area Above the Ordinal Dominance Graph and the Area Below the Receiver Operating Characteristic Graph. Journal of Mathematical Psychology, 12(4), 387-415 (1975).
2. E. Hussain; The ROC Curve Model from Generalized Exponential Distribution. Pakistan Journal of Statistics and Operations Research, 7(2), 323-330 (2011).
3. J.A. Hanley and B.J.M. Neil; The Meaning and use of the Area under a Receiver Operating Characteristic (ROC) Curve. Radiology, 143, 29-36 (1982).
4. R.D. Gupta and D. Kundu; Generalized Exponential Distribution. The Australian and Newzealand Journal of Statistics, 41, 173-188 (1999).
5. R.D. Gupta and D. Kundu; Generalized Exponential Distribution: Existing Results and Some Recent Developments. Journal of Statistical Planning and Inference, 136, 3130-3144 (2007).
6. R.V. Vardhan; S. Pundir and G. Sameera; Estimation of Area under the ROC Curve using Exponential and Weibull Distributions. Bonfring International Journal of Data Mining, 2(2), 52-56 (2012).
7. S. Balaswamy and R.V. Vardhan; A Class of Hybrid ROC Curves for Skewed Distributions. Lap Lambert Academic Publishing, (2023).
8. S. Balaswamy and R.V. Vardhan; An Anthology of Parametric ROC Models. Research & Reviews: Journal of Statistics, 5(2), 32-46 (2016).
9. S. Balaswamy and R.V. Vardhan; Estimating the Area under the ROC Curve in the Framework of Lindley Centered Distributions. Statistics and Applications, 19(2), 231 – 240 (2021).
10. S. Balaswamy and R.V. Vardhan; Interval Estimation and Inferential Procedures of the Measures of Hybrid ROC (HROC) Curve. Biostatistics and Biometrics Open Access Journal, 3(4), 1-10 (2017).
11. W.J. Krzanowski and D.J. Hand; ROC Curves for Continuous Data Monographs on Statistics and Applied Probability. CRC Press, Taylor and Francis Group, New York, (2009).

ISSN(P) 2350-0174

ISSN(O) 2456-2378

Journal Content