Features

BERKELEY, July 7 -- Leo Breiman, professor emeritus of statistics at the University of California, Berkeley, and a man who loved to turn numbers into practical and useful applications, died July 5 at his Berkeley home after a long battle with cancer. He was 77.

A faculty member in the Department of Statistics since 1980, Breiman was a man of remarkably wide-ranging interests, including probability theory, mathematical statistics and cutting-edge statistical computation. He inspired and challenged his students and once said he'd advise a student considering a career in statistics today to "remember that the great adventure of statistics is in gathering and using data to solve interesting and important real world problems."

Late last month, Breiman was awarded the SIGKDD Data Mining and Knowledge Discovery Innovation Award, the highest technical award in these two areas. It is given for significant technical innovations in data mining and knowledge discovery that have been transferred to practical uses in significant ways or have notably influenced the direction of research and development in the field.

...

In particular," said Bickel, "Breiman developed one of the most successful state-of-the-art classification programs, 'Random Forest.' This method was based on a series of new ideas that he developed in papers during the last seven years, and it is extensively used in government and industry."

Breiman's best known work is considered to be "Classification and Regression Trees," a work in collaboration with three other scholars that facilitates practical applications, such as the diagnosis of diseases, from a multitude of symptoms.

During his career, Breiman was elected a member of the National Academy of Sciences and of the American Academy of Arts and Sciences.

He is survived by his wife, Mary Lou, of Berkeley; and daughters, Rebekah and Jessie, both of Salt Lake City.

Here is the rest of the obituary.