Richard Olshen, statistician who created groundbreaking machine learning applications, dies at 81

The Stanford Medicine professor was best known for his work in recursive partitioning, an aspect of machine learning.

- By Jennifer Welsh

Richard Olshen

Richard Allen Olshen, PhD, emeritus professor of biomedical data science at the Stanford School of Medicine, died Nov. 8 at the age of 81.

Chief of the biostatistics division in the Department of Health Research and Policy from 1998 until 2016, when he joined the Department of Biomedical Data Science, Olshen focused his research on statistics and their applications to medicine and biology.

“Richard Olshen had an unyielding passion for academia and a profound love for Stanford University,” said Lloyd Minor, MD, dean of the School of Medicine and vice president for medical affairs at Stanford University. “He dedicated his career to research in biomedical statistics, and so many of the advances we see today stand on the shoulders of his pioneering work.”

The statistical method he co-developed, known as classification and regression trees (CART), led to many advancements. He started working on CART in 1974 as a member of the Computation Research Group at the Stanford Linear Accelerator Center. CART models use a computer-based decision tree to predict a specific outcome based on different input values. These decision trees are used to classify patients and predict various outcomes. His well-known book Classification and Regression Trees described the motivation, algorithms and mathematical theory of CART.

“His work on CART had a big effect and was ahead of its time,” said Bradley Efron, PhD, an emeritus professor of statistics and biomedical data science. “It’s the kind of data analytic device that was very handy and is used thousands of times across different fields.”

Olshen applied statistics and mathematics to medicine and biology. Subjects he worked on included how people walk, the compression and analysis of medical images, and the genetics of HIV and hypertension. As a part of his research into how people walk, he coauthored The Development of Mature Walking.

As division chief, Olshen was a leader in the shaping of biostatistics at Stanford and was a key figure in the development of the biostatistics workshop and data studio.

“He was a founding figure of the biostatistical group,” said Lu Tian, ScD, a professor of biomedical data science. “His legacy will be felt well beyond Stanford — he will be known to many future generations of statisticians.”

He also served as a mentor, going out of his way to help others and ensure they were successful in their careers, his colleagues said. He supported those junior to him and was generous with his time, Tian said.

Olshen’s colleagues said he had a remarkable ability to tell it like it is. He held himself and those around him to a high standard, they said, adding that he was a walking encyclopedia of mathematical statistics and sports trivia who was warm and supportive with a keen sense of humor.

Richard had a great personality; it was really fun to be around him. He’s the kind of person who’s really missed.

“Richard had a great personality; it was really fun to be around him. He’s the kind of person who’s really missed,” Efron said. “When I saw him coming down the hall, I was happy because it was always fun to talk to him. He was very popular, really a warm guy and a great gossip. And his gossip was usually accurate because he knew everybody.”

Olshen was known for asking tough and long questions. “The trick was not rushing to answer him,” Tian said. “Instead, if you waited until the end, Richard would often deliver the best answers himself.”

Early life and career

Born in Portland, Oregon, on May 17, 1942, Olshen’s father worked as an actuary, while his mother was a homemaker. He spent his early years in Chevy Chase, Maryland, but lived most of his life in California.

After graduating from Burlingame (California) High School in 1959, he followed his future wife, Vivian, to the University of California, Berkeley. He earned a degree in statistics in 1963.

He and Vivian married before heading off to Yale University for his graduate work in statistics. He earned his PhD in 1966 and stayed to serve as a research staff statistician and lecturer through 1967.

In 1967 he returned to California as an assistant professor in the statistics department at Stanford University, the beginning of his long connection with the university. His son, Adam, was born at Stanford Hospital in 1968.

The young family briefly moved to New York City after he accepted a role as a visiting professor at Columbia University before settling into a tenured faculty position as associate professor of statistics and mathematics at the University of Michigan in 1972. His daughter, Elyse, was born in Michigan in 1972.

Olshen’s wife died in 1973, which led him to return to California for support from his family as a newly single parent. He returned to working at Stanford University.

In 1975, Olshen moved to the mathematics department at UC San Diego, where he earned the rank of full professor in 1977. He helped found what became the Moores Cancer Center at UC San Diego Health.

Richard Olshen, with his wife, Susan Olshen, and two of their grandchildren.
Courtesy of Manisha Desai

In 1977, he met his second wife, Susan, through a mutual friend. The two married in La Jolla, California, on July 1, 1979. With her children, Steve and Rachel, they became a family of six.

After a decade in southern California, he returned to Stanford in 1989 as a professor of biostatistics. He remained at Stanford until his retirement in 2016 at the age of 74, though he continued to mentor others and work on a molecular genetics project with Tian. He returned to campus to participate in seminars, interact with colleagues and collaborate on research papers.

A sports nut and family man

Olshen’s family described him as a dedicated husband, nurturing father and doting grandfather. He loved watching and discussing sports and attended his children’s and grandchildren’s soccer and basketball games, along with their recitals and concerts.

He was happiest when the family was together, his family said, though he also enjoyed walking along the beach and traveling the world with his wife.

He was also an active member of his synagogue and read the weekly Torah portion.

“Dad taught me to be tough in the face of adversity — and in the value of hard work,” said his daughter, Elyse Kharbanda, MD. “Dad was someone I could call on for advice and support — I valued his wisdom and will miss that.”

Olshen was an elected fellow of the Institute of Mathematical Statistics, the American Association for the Advancement of Science, the American Statistical Association and the Institute of Electrical and Electronics Engineers. He was a former Guggenheim fellow and research scholar at the American Cancer Society.

He is survived by his wife, Susan; his four children; eight grandchildren; and a sister.

About Stanford Medicine

Stanford Medicine is an integrated academic health system comprising the Stanford School of Medicine and adult and pediatric health care delivery systems. Together, they harness the full potential of biomedicine through collaborative research, education and clinical care for patients. For more information, please visit med.stanford.edu.

2024 ISSUE 1

Psychiatry’s new frontiers