Extreme value analysis helps predict how likely extreme events are to occur based on past data. It focuses on rare but severe events, like high or low values in a process. Researchers use methods like period maxima and GEV distribution to identify extreme values. Different distributions like Gumbel and Frechet can also be used. The Gumbel distribution is a special case of the Hosking distribution. Extreme value max distribution is for extreme maxima, while extreme value min distribution is for extreme minima.