Abstract:This paper investigates stochastic finite matrices and the corresponding finite Markov chains constructed using recurrence matrices for general families of orthogonal polynomials and multiple orthogonal polynomials. The paper explores the spectral theory of transition matrices, utilizing both orthogonal and multiple orthogonal polynomials. Several properties are derived, including classes, periodicity, recurrence, stationary states, ergodicity, expected recurrence times, time-reversed chains, and reversibility. Furthermore, the paper uncovers factorization in terms of pure birth and pure death processes. The case study focuses on hypergeometric orthogonal polynomials, where all the computations can be carried out effectively. Particularly within the Askey scheme, all descendants under Hahn (excluding Bessel), such as Hahn, Jacobi, Meixner, Kravchuk, Laguerre, Charlier, and Hermite, present interesting examples of recurrent reversible birth and death finite Markov chains. Additionally, the paper considers multiple orthogonal polynomials, including multiple Hahn, Jacobi-Piñeiro, Laguerre of the first kind, and Meixner of the second kind, along with their hypergeometric representations and derives the corresponding recurrent finite Markov chains and time-reversed chains.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to explore the connection between finite homogeneous Markov chains and sequences of orthogonal polynomials, especially the zeros corresponding to these polynomials and the truncation of the recurrence matrix. Specifically, the authors extend the ideas in previous studies and focus on finite Markov chains rather than countable Markov chains. In the context of finite Markov chains, the role of 1 is replaced by the largest zero of the orthogonal polynomial \(P_n(x)\). By using the zeros of orthogonal or multiple - orthogonal polynomials (in the case of multiple - orthogonal polynomials, also using the determinants of type - I multiple - orthogonal polynomials), the authors provide a step - by - step method to construct stochastic matrices based on these polynomials. These matrices are associated with birth - death Markov chains in the case of orthogonal polynomials, and with Markov chains beyond the birth - death range in the case of multiple - orthogonal polynomials.
The paper also explores various properties of these finite Markov chains, including recurrence or transience of states, periodicity, ergodicity, steady - state, expected return time, reversibility and reverse chains. In addition, the paper describes a process of decomposing a stochastic matrix into bidiagonal stochastic matrices, each of which simulates a pure birth or pure death Markov chain. This decomposition is particularly relevant for constructing the corresponding urn models.
Overall, the goal of the paper is to construct and analyze finite Markov chains through the theory of orthogonal polynomials, especially polynomial families in the Askey scheme, such as Hahn, Jacobi, Meixner, Kravchuk, Laguerre, Charlier and Hermite, etc., and their hypergeometric representations. At the same time, the paper also provides two Mathematica codes for readers to generate various probability elements of Markov chains under different parameter sets, such as steady - state, expected return time and pure birth / pure death stochastic factorizations.