发帖
雷达卡
Fluctuation Credibility: 90% of the time, total claim between 90% to 110% of the expected claims. CV of frequency and severity are known.Greenwood approximation of K-M estimator.
Frequency is 0,1,2, with probability 0.6,0.15.0.25,severity is exponential with theta=1, what's the probability that aggregate claim >1.2
1 question with non-parametric bayes model: 2 policyholders, 4 year, different claim number in different years, what's the bayes estimate for policyholder 1 in year 5?
f(x)=p(x)*e^(r(x))/q(x), where q(x)=1/a*exp(-x/a), where p(x) doesn't depend on a, and q(x) is a normalizing function(or something like that), E(X)=a+1, what's var(x)?
An insured has a poisson distribuiton, where lambda's distribution is exponential with mean = 5, first 5 year total claim is 50, what's the bayes estimate of the cliam in year 6?
lognormal distribution, with u=5, sigma=2, loss inflate 7% each year, what's the 2nd raw moment in year 3.
frequency is poisson with mean =2, severity is exponential with mean = 300( or some other number), deductable = 250.Inversion method to simulate both freq and severity: 0.5 for freq, 0.99. .75 and 0.6 for severity, what's the aggregate claim?
likely hood ratio test, LN(sum(Xi+2000))=100(i=1...100)1 kernel question
京公网安备 11010802022788号
