Andrei Nikolaevich Kolmogorov (1903-1987)

Andrei Nikolaevich Kolmogorov was born in Tambov, halfway between Moscow and Volgograd. His mom passed away during childbirth, so he was raised by his aunts in a small village in what is now southern Russia. As a child, he posed math problems to his classmates. In secondary school, he learned higher mathematics from an encyclopedia. For a physics project, he created plans for an impossible perpetual motion machine by concealing errors in his work. When he applied early for graduation, the committee granted him the certificate without giving him the required examination, and he was sorely disappointed because he had been looking forward to the challenge. He enrolled in Moscow State University at age 17 studying mathematics and later joined the faculty.

Up until his time, people had used probability almost exclusively to answer specific questions without developing a formalized theory of probability. Probability theory needed to include internally consistent axioms and theorems. Kolmogorov provided the groundwork for probability theory, writing the axioms for probability that are used to teach probability now. To formalize the axioms, he combined set theory and measure theory with probability and used Venn diagrams. The axioms allowed people to study probability more abstractly.

Even though Kolmogorov and his students contributed to the development of probability and statistical theory, the Soviet Union's official stance was that the government had all the correct answers, so any competing ideas were rejected. The term for random variable in Russian translates to accidental magnitude. According to Communist theory, there are no accidents; everything is planned. So, the use of mathematical statistics dwindled during Communist control. This extended beyond statistics to other sciences. For example, Stalin supported a biologist named Trofim D. Lysenko who claimed that plant and animal traits come only from the environment, which was controllable rather than inheritance, which involves chance processes.

While best known for the probability axioms, he also considered the questions of how to work with data that is affected by previous data, such as data Ronald Fisher (1890-1962) recorded consisting of weekly weights of his newborn son. Fisher observed that his son's weight one week would depend on his weight the previous week, and even the amount of change could depend on the previous week. Kolmogorov labelled this idea a stochastic process and laid foundations for further study.