M. Mohammadi,M. Hashempour,O. Kamari
Abstract:In this paper, we introduce a novel way to quantify the remaining inaccuracy of order statistics by utilizing the concept of extropy. We explore various properties and characteristics of this new measure. Additionally, we expand the notion of inaccuracy for ordered random variables to a dynamic version and demonstrate that this dynamic information measure provides a unique determination of the distribution function. Moreover, we investigate specific lifetime distributions by analyzing the residual inaccuracy of the first-order statistics. Nonparametric kernel estimation of the proposed measure is suggested. Simulation results show that the kernel estimator with bandwidth selection using the cross-validation method has the best performance. Finally, an application of the proposed measure on the model selection is provided.
engineering, industrial,operations research & management science,statistics & probability
What problem does this paper attempt to address?
### Problems Addressed by the Paper
This paper aims to introduce a new method to quantify the residual inaccuracy of order statistics using **residual extropy**. Specifically, the authors explore various properties and characteristics of this new measure and extend it to a dynamic version. By analyzing the residual inaccuracy of the first order statistic, the study investigates specific lifetime distributions. Additionally, the paper proposes a non-parametric kernel estimation method and validates its effectiveness through simulation experiments. Finally, the paper discusses the application of the proposed measure in model selection.
### Main Content
1. **Introduction**:
- Introduces the basic concepts of non-negative continuous random variables and their distribution functions, probability density functions, and survival functions.
- Discusses the applications of order statistics in fields such as material strength, statistical inference, reliability theory, goodness-of-fit tests, quality control, outlier detection, and characterization of probability distributions.
- Mentions Shannon's information measure and Kerridge's inaccuracy measure, and discusses their importance in information theory.
2. **Dynamic Residual Inaccuracy Measure**:
- Introduces a dynamic residual inaccuracy measure based on extropy.
- Defines a dynamic residual inaccuracy measure between two residual lifetime distributions.
- Explores the properties of these measures, including boundaries and symmetry.
3. **Analysis of Residual Inaccuracy**:
- Studies the residual inaccuracy of order statistics and its significance in characterization results.
- Analyzes the residual inaccuracy of specific lifetime distributions through concrete examples.
4. **Non-Parametric Estimation**:
- Proposes a non-parametric kernel estimation method to estimate the proposed measure.
- Validates the performance of the kernel estimator in bandwidth selection through simulation experiments, finding that the kernel estimator performs best when using the cross-validation method for bandwidth selection.
5. **Practical Application**:
- Demonstrates the behavior of the estimator in practical applications using real datasets.
### Summary
This paper provides a new method for evaluating the inaccuracy of order statistics by introducing a dynamic residual inaccuracy measure based on extropy. This method not only offers theoretical innovation but also performs well in practical applications, particularly in model selection and lifetime data analysis.