Algorithm, probability, and prime numbers

Abstract

The results of the study provide guidelines for the development and applications of algorithms. When the number of steps for calculating an assumption tends to infinity, probability theory can be applied to predict whether the assumption holds or not. This characteristic is related to the number of steps for verifying arbitrarily large prime numbers. This study proved that $\pi (n)-Li(n)=o(M(n)\sqrt{Li(n)})$ almost certainly holds without any assumptions. Here, $\pi (n)$ is the number of primes not greater than $n$, $Li(n)$ is a logarithmic integral function, and $M(n)$ is an arbitrary function such that $M(n)\rightarrow\infty$. This result implies that the Riemann hypothesis holds as the falseness of the Riemann hypothesis leads to a contradiction.