13_AS_2_lec_a

# Or integrated hazard t t s ds 0 mj j tj t since

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Maximum Likelihood with censoring The likelihood as a function of the discrete hazard function Kaplan-Meier (product limit) estimator Nelson-Aalen estimator 4 Comparing survival functions Introductory example A hypothesis test Special cases 26/42 Actuarial Statistics – Module 2: Non-parametric methods Estimating the lifetime distribution: non-parametric approach Nelson-Aalen estimator ˆ Another non-parametric approach for F (t ) is the Nelson-Aalen estimator (from Nelson, 1971 and Aalen, 1978), which is based on the cumulative hazard (or integrated hazard) t Λt = µs ds + 0 mj . j :tj ≤t ˆ Since in our setting we do not have continuous increases in F (t ) (only jumps – see previous sub-section) we focus on the second half only and use the ML estimator for λj to approximate the mj ’s such that dj ˆ Λt = nj j :tj ≤t Finally, ˆ ˆ S ( t ) = e − Λt . 26/42 Actuarial Statistics – Module 2: Non-parametric methods Estimating the lifetime distribution: non-parametric approach Nelson-Aalen estimator Variance of Nelson-Aalen estimator ˜ Let Λ (t ) denote the estimatOR. Its variance is approximated as (p. 33 of yellow book): ˜ var Λ (t ) ≈ j :tj ≤t dj (nj − dj ) nj3 Example Consider the following recorded data for the lifetime of a group of rats 2, 4+ , 5, 8, 12, 12+ , 17, 18+ , 18+ , 18+ , where + denotes the censored. Derive 95% conﬁdence intervals for ˆ the N-A estimators Λ(5). 27/42 Actuarial Statistics – Module 2: Non-parametric methods Estimating the lifetime distribution: non-parametric approach Nelson-Aalen estimator Relationship between the Kaplan-Meier and Nelson-Aalen Denote: ˆ the Kaplan-Meier estimate of the survival function by SKM (t ) ˆ the Nelson-Aslen estimate of the survival function by SNA (t ) Then ˆ SKM (t ) = 1− j :tj ≤t dj nj ≈ exp − j :tj ≤t (using e −x ≈ 1 − x for small |x |) ∧ dj = exp −Λt nj ˆ = SNA (t ) Furthermore, ˆ ˆ SKM (t ) < SNA (t ) 28/42 for t1 ≤ t ≤ tmax . (remember ln(1 − x ) < −x for 0 < x < 1) Actuarial Statistics – Module 2: Non-parametric methods Comparing survival functions 1 Introduction 2 Censoring and truncation Censoring Truncation Likelihood for censored and truncated data 3 Estimating the lifetime distribution: non-parametric approach Preliminaries Maximum Likelihood with censoring The likelihood as a function of the discrete hazard function Kaplan-Meier (product limit) estimator Nelson-Aalen estimator 4 Comparing survival fu...
View Full Document

## This document was uploaded on 04/03/2014.

Ask a homework question - tutors are online