Abstract:Three versions of the Weak Law of Large Numbers are proposed for weakly dependent and generally speaking non-equally distributed random variables, with finite or possibly infinite expectations.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the application and extension of the Weak Law of Large Numbers (WLLN) under specific conditions. Specifically, the author proposes a new version of the Weak Law of Large Numbers for sequences of random variables that are not independent and may have different distributions. These conditions include, but are not limited to:
1. **Uniform integrability**: For a sequence of random variables with finite expected values, their uniform integrability in some form (such as uniform integrability in the sense of Cesàro means) is required. This helps to deal with situations where although the expected value of each random variable exists, they may not be completely independent.
2. **Weak - dependence conditions**: New weak - dependence conditions are introduced. These conditions allow for a certain degree of dependence between random variables, but this dependence must be weak enough to ensure that the law of large numbers still holds. For example, the condition $\frac{1}{n}\sum_{k = 1}^{n}E|E(\xi_{k}|\xi_{1}+\ldots+\xi_{k - 1})|\to0$ is used as a measure of weak - dependence.
3. **The case without finite expected values**: The paper also explores the law of large numbers when random variables do not have finite expected values. In this case, a concept similar to "principal value" is used to define the expected value, and specific conditions (such as $\frac{\tilde{a}_{n}}{n}\to0$) are used to ensure the validity of the law of large numbers.
### Specific problem description
1. **The case of finite expected values**:
- **Theorem 1**: If the sequence of random variables $(\xi_{n})$ is uniformly integrable in the sense of Cesàro means and the expected value of each random variable is zero, then $E\left|\frac{S_{n}}{n}-\frac{a_{n}}{n}\right|\to0$, where $S_{n}=\sum_{k = 1}^{n}\xi_{k}$ and $a_{n}=\sum_{k = 1}^{n}E(\xi_{k}|\xi_{1},\ldots,\xi_{k - 1})$. Further, if $\frac{a_{n}}{n}\to0$ (which is a condition of weak - dependence), then $\frac{S_{n}}{n}\to0$.
2. **Weak - dependence conditions**:
- **Theorem 2**: If the sequence of random variables $(\xi_{n})$ is uniformly integrable in the sense of Cesàro means and satisfies the weak - dependence condition $\frac{1}{n}\sum_{k = 1}^{n}E|E(\xi_{k}|\xi_{1}+\ldots+\xi_{k - 1})|\to0$, then $\frac{S_{n}}{n}\to0$.
3. **The case without finite expected values**:
- **Theorem 3**: If the sequence of random variables $(\xi_{n})$ is uniformly integrable in the sense of Cesàro means and satisfies the uniform integrability and convergence conditions of certain functions, as well as the weak - dependence condition $\frac{\tilde{a}_{n}}{n}\to0$, then $\frac{S_{n}}{n}\to0$.
### Summary
The main contribution of the paper is to extend the existing law of large numbers theory, making it applicable to a wider range of sequences of random variables, especially those that are not completely independent and may have different distributions. By introducing new weak - dependence conditions and uniform integrability conditions, the author provides a rigorous mathematical proof for the law of large numbers in these complex situations. These results have important theoretical and applied values in probability theory and statistics.