경쟁 위험 모델에서 역확률 중도 절단 가중치를 이용한 시간 가변 AUC 추정량 비교 / Comparing time-dependent AUC with inverse probability of censoring weighting estimators in competing risk model
- Resource Type
- Dissertation/ Thesis
- Authors
- 성재현 / Seong Jae-hyun
- Source
- Subject
- time-dependent AUC
competing risk
inverse probability of censoring weighting
survival analysis
censoring
- Language
- Korean
The time-dependent ROC curve have been used to measure model parameters in survival analysis. The curve is defined using cumulative sensitivity and dynamic specificity that is used to time-dependent variable. With competing risk model, we define AUC that classify the case and the control by considering “Competing events”. Blanche(2013) attempted to estimate time-dependent AUC by applying the inverse probability of censoring weighting(IPCW) estimator. This simulation aims to compare performance by expanding the conditions in Blanche(2013)’s simulation.