B1538
Title: Nonparametric estimation of the cross-ratio function with splines
Authors: Marsha Nugroho - Hasselt University (Belgium) [presenting]
Steven Abrams - University of Antwerp and UHasselt (Belgium)
Anneleen Verhasselt - Hasselt University (Belgium)
Abstract: In the analysis of bivariate time-to-event data, measuring the strength of the association between the event times is of interest. Several global association measures like Kendall's tau, Pearson correlation and odds ratio exist. However, the interest is in local association measures, more specifically, the cross-ratio function. A flexible nonparametric estimator is proposed based on splines for the cross-ratio function, as such extending the parametric polynomial estimator in the context of right censored data. Moreover, a P-spline penalty is added to impose more smoothness. The finite sample performance of the estimator of the cross-ratio function is illustrated on real and simulated data.
