Extreme Events and Extreme Value Theory (EVT) [27]

Go to: Summary | Previous | Next   
Bullet points include: Let N(u) be number of excesses Yi over the threshold u in the sample. We assume that the excesses Y1, ..., YN(u) are distributed according to a GPD. Set. Then ML-estimates are found by solving (c.f. ‘Hill’ estimator):

Contents | Prev | Next | ERM Lecture Series

Desktop view | Switch to Mobile