Abstract:The purpose of this article is to show that Theorems 2.2-2.5 from [1] apply to the product of random matrices considered by Grama, Le Page, and Peigné [2]. This allows us, in particular, to emphasize the general nature of the formulation of our theorems in [1] by showing that our assumptions are verified for previous models.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to show how to construct a suitable Banach space for the product of random matrices to meet specific assumptions (M1 - M5), so as to ensure that the previously proposed theorems (2.2 - 2.5) are applicable to the case of the product of random matrices. Specifically, the paper addresses the following points:
1. **Adapting to Existing Assumptions**: The existing Banach space defined in [2] does not fully meet the assumptions M1 - M5. In particular, in assumption M4, it is required that the function \( f \) be bounded by a function in the Banach space. However, the Banach space given in [2] is contained in the set of bounded functions, and the function \( f=\rho \) is not bounded. Therefore, it is necessary to modify the existing Banach space to ensure that it meets all the assumptions.
2. **Constructing a New Banach Space**: The paper introduces a new Banach space \( B \) and defines the norm \( \|h\|_B \) on it, so that this new space can better adapt to the characteristics of the product of random matrices and meet the assumptions M1 - M5. Specifically, the new space \( B \) contains continuous functions \( h \), and the norms of these functions are composed of three parts:
\[
\|h\|_B := |h|_\theta + k_{\varepsilon, \alpha}(h) + k'_{\varepsilon, \beta}(h),
\]
where,
\[
|h|_\theta = \sup_{(g,u) \in X} \frac{|h(g,u)|}{N(g)^\theta},
\]
\[
k_{\varepsilon, \alpha}(h) = \sup_{g \in G} \sup_{u \neq v} \frac{|h(g,u) - h(g,v)|}{d(u,v)^\varepsilon N(g)^\alpha},
\]
\[
k'_{\varepsilon, \beta}(h) = \sup_{g \neq g'} \sup_{u \in P(\mathbb{R}^d)} \frac{|h(g,u) - h(g',u)|}{\|g - g'\|^\varepsilon N(g)^\beta N(g')^\beta}.
\]
3. **Verifying the Assumption Conditions**: The paper further verifies that the newly constructed Banach space \( B \) indeed meets the assumption conditions M1 - M5, and shows how to use these assumption conditions to prove several important results regarding the product of random matrices, such as the existence of spectral gaps, the central limit theorem, etc.
Through these steps, the paper successfully applies the theorems previously proven in the general case to the special case of the product of random matrices, thus providing a solid theoretical basis for the study of the product of random matrices.