On One Extension of a Martingale
B Gnedenko
Abstract
B Gnedenko
Abstract
In this paper we introduce an ε‐martingale and a strongε‐martingale. The first is defined by the inequality $|{\bf E}(X_t\,|\,{\cal F}_s)- X_s|\le \varepsilon$, and the second one can be obtained from the ε‐martingale by replacing in the definition fixed time moments with stopping times. The paper proves that a right‐continuous ε‐martingale is a strong 2ε‐martingale. At the same time we construct an example of a right‐continuous ε‐martingale which is not a strong ε‐martingale for any $a<2$. We show that the dependence between ε‐martingales and strong ε‐martingales has no analogues for ε‐submartingales. We also give the criterion for testing if a right‐continuous with left limits process is a strong ε‐martingale or not. The criterion is based on the possibility of uniform approximation of the process by a martingale with precision $\varepsilon/2$.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In this paper we introduce an ε‐martingale and a strongε‐martingale. The first is defined by the inequality $|{\bf E}(X_t\,|\,{\cal F}_s)- X_s|\le \varepsilon$, and the second one can be obtained from the ε‐martingale by replacing in the definition fixed time moments with stopping times. The paper proves that a right‐continuous ε‐martingale is a strong 2ε‐martingale. At the same time we construct an example of a right‐continuous ε‐martingale which is not a strong ε‐martingale for any $a<2$. We show that the dependence between ε‐martingales and strong ε‐martingales has no analogues for ε‐submartingales. We also give the criterion for testing if a right‐continuous with left limits process is a strong ε‐martingale or not. The criterion is based on the possibility of uniform approximation of the process by a martingale with precision $\varepsilon/2$.
Key concepts: Martingale (probability theory), Doob's martingale inequality, Mathematics, Local martingale, Martingale difference sequence, Pure mathematics, Applied mathematics