One-dimensional Stein's method with bespoke derivatives

Gilles Germain,Yvik Swan

2023-11-03

Abstract:We introduce a version of Stein's method of comparison of operators specifically tailored to the problem of bounding the Wasserstein-1 distance between continuous and discrete distributions on the real line. Our approach rests on a new family of weighted discrete derivative operators, which we call bespoke derivatives. We also propose new bounds on the derivatives of the solutions of Stein equations for Integrated Pearson random variables; this is a crucial step in Stein's method. We apply our result to several examples, including the Central Limit Theorem, Polya-Eggenberger urn models, the empirical distribution of the ground state of a many-interacting-worlds harmonic oscillator, the stationary distribution for the number of genes in the Moran model, and the stationary distribution of the Erlang-C system. Whenever our bounds can be compared with bounds from the literature, our constants are sharper.

Probability

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to use the Stein method to quantify the Wasserstein - 1 distance (also known as the Kantorovitch or L1 distance) between continuous distributions and discrete distributions in one - dimensional space. Specifically, the author introduces a new family of weighted discrete derivative operators, called "bespoke derivatives", to improve the effect of the Stein method when comparing these distributions. The main contribution of the paper is to provide a general abstract bound (see Theorem 1.1), which enables the handling of multiple specific examples, such as the central limit theorem, the Polya - Eggenberger urn model, the empirical distribution of the ground state of multi - interacting - world harmonic oscillators, the steady - state distribution of the number of genes in the Moran model, and the steady - state distribution of the Erlang - C system. Through this method, the author can obtain more precise constant bounds than those in the existing literature.

The key points of the paper include:
- **Problem Definition**: Quantify the one - dimensional Wasserstein distance \(W_1(X, Z)\) between the continuous random variable \(Z\) and the discrete random variable \(X\).
- **Method Innovation**: Introduce a new weighted discrete derivative operator (bespoke derivatives) and propose a new derivative bound for the solution of the Stein equation.
- **Theoretical Contribution**: Provide a general abstract bound theorem (Theorem 1.1) applicable to the comparison of multiple distributions.
- **Application Examples**: Demonstrate the effectiveness of the method through specific application examples, including but not limited to the central limit theorem, the Polya - Eggenberger urn model, etc.

Through these contributions, the paper provides new tools and perspectives for the Stein method in dealing with the distance problem between continuous and discrete distributions.