Yuan-Shih Chow (1924–2022): The Architect of Optimal Stopping
Yuan-Shih Chow was a titan of 20th-century probability theory and mathematical statistics. Over a career spanning seven decades, he transformed the way mathematicians understand "stopping rules"—the logic of when to cease collecting data and make a decision. His work bridged the gap between abstract measure theory and the practicalities of sequential analysis, leaving an indelible mark on finance, engineering, and clinical trials.
1. Biography: From War-Torn China to Columbia University
Yuan-Shih Chow was born on September 1, 1924, in Xiangyang, Hubei Province, China. His early education was disrupted by the Second Sino-Japanese War, forcing him to move frequently to escape conflict. Despite these hardships, he showed an early aptitude for mathematics, eventually enrolling at the National Chekiang (Zhejiang) University, where he studied under the legendary geometer Su Buqing. He graduated with a B.S. in 1947.
Following the Chinese Civil War, Chow moved to Taiwan in 1949, where he served as an assistant at National Taiwan University. Seeking advanced training, he emigrated to the United States in 1954 to attend the University of Illinois at Urbana-Champaign. There, he studied under David Blackwell, one of the most influential statisticians of the era. Chow earned his Ph.D. in 1958 with a dissertation focused on martingale theory.
After a brief stint at IBM Research and an assistant professorship at Purdue University, Chow joined the faculty of Columbia University in 1968. He remained at Columbia for the rest of his career, eventually becoming Professor Emeritus. He passed away on March 3, 2022, at the age of 97.
2. Major Contributions: Martingales and Optimal Stopping
Chow’s intellectual legacy is defined by his rigorous development of two interconnected fields:
Optimal Stopping Theory
Chow is perhaps best known for formalizing the theory of Optimal Stopping. This involves choosing a time to take a particular action (like selling a stock or ending a clinical trial) to maximize an expected reward. Before Chow, many of these problems were solved piecemeal; he provided the overarching mathematical framework that unified them.
Martingale Theory
A "martingale" is a model of a fair game where the future expected value is equal to the current value, regardless of past events. Chow developed fundamental Martingale Convergence Theorems and inequalities. His work showed how martingales could be used to solve complex problems in sequential analysis—where the number of observations is not fixed in advance but depends on the data already collected.
The Chow-Robbins Game
In collaboration with Herbert Robbins, he developed what is now known as the Chow-Robbins Game. This is a classic problem in optimal stopping: if you flip a coin and receive a reward based on the proportion of heads, when is the mathematically "perfect" time to stop? Their work proved that an optimal strategy exists, though it is notoriously difficult to calculate, sparking decades of further research.
3. Notable Publications
Chow’s bibliography includes several texts that remain cornerstones of graduate-level mathematics:
- Great Expectations: The Theory of Optimal Stopping (1971): Co-authored with Herbert Robbins and David Siegmund. This is widely considered the "bible" of optimal stopping theory, providing the first comprehensive treatment of the subject.
- Probability Theory: Independence, Interchangeability, Martingales (1978): A rigorous textbook (co-authored with Henry Teicher) that became a standard reference for doctoral students in probability.
- Admissible Stopping Rules for Sn/n (1967): A seminal paper in the Annals of Mathematical Statistics that laid the groundwork for the Chow-Robbins game.
- Sample Size Determination (1991): A practical guide for statisticians on how to calculate the necessary size of a study to ensure valid results.
4. Awards and Recognition
Chow’s contributions earned him the highest honors in his field:
- Academician of Academia Sinica (1974): Elected to Taiwan’s national academy, reflecting his status as one of the world's leading Chinese mathematicians.
- Fellow of the Institute of Mathematical Statistics (IMS): A distinction reserved for those who have demonstrated sustained excellence in statistical research.
- Director of the Institute of Mathematics, Academia Sinica (1970–1977): During this period, he took a leave from Columbia to return to Taiwan, where he played a pivotal role in modernizing the nation's mathematical research infrastructure.
5. Impact and Legacy
Chow’s work provided the mathematical "scaffolding" for modern Financial Mathematics. The Black-Scholes model and other option-pricing theories rely heavily on the martingale and optimal stopping theories that Chow refined.
Beyond the equations, his legacy lives on through the "Chow School" of statisticians. He supervised over 30 Ph.D. students at Columbia, many of whom, such as Tze Leung Lai, became world-renowned scholars in their own right. He is remembered for his "Hubei temperament"—a combination of fierce intellectual rigor and a deeply kind, mentoring spirit.
6. Collaborations
Chow was a highly collaborative researcher, often working at the intersection of different mathematical philosophies.
- Herbert Robbins: His most frequent and famous collaborator. Together at Columbia, they formed a powerhouse duo that defined the "Golden Age" of sequential analysis.
- David Blackwell: His mentor, who introduced him to the Bayesian perspective and the elegance of martingale theory.
- David Siegmund: A younger colleague with whom he co-authored Great Expectations, bridging the gap between foundational theory and modern applications.
7. Lesser-Known Facts
- The "Secretary Problem": While Chow didn't invent the "Secretary Problem" (how many candidates to interview before picking the best one), his work on optimal stopping provided the definitive proof for the 1/e (roughly 37%) rule, which states you should reject the first 37% of candidates and then pick the next one who is better than all previous ones.
- A Bridge to China: Following the normalization of US-China relations in the 1970s, Chow was one of the first prominent mathematicians to return to mainland China to deliver lectures, helping to reintegrate Chinese mathematics into the global community.
- Late Career Vitality: Chow remained mathematically active well into his 80s, often seen in the Columbia mathematics lounge discussing the latest proofs with students a third of his age.