On Ordered Fuzzy Numbers Generated By Time Series

TL;DR

This paper introduces a novel trapezoidal ordered fuzzy number representation for windowed time series, inspired by Japanese candlesticks, and demonstrates its application using financial data from automotive companies.

Contribution

It presents a new fuzzy number model for time series analysis based on candlestick concepts, enhancing descriptive statistics and interpretability.

Findings

New fuzzy number representation for time series

Application to automotive financial data

Potential for broader data-driven insights

Abstract

This paper proposes a new trapezoidal ordered fuzzy number representation of windowed time series based on the idea of a Japanese candlestick to those previously proposed in the literature. We define and illustrate several descriptive statistics based on the information contained in the ordered fuzzy numbers. We utilize financial trading data from three automotive companies as a case study. Further expansion can be applied to any other forms of time series data to offer further insight into many other data driven situations.