A new visualization method for pairwise time-series data with random walk plot
Sangmin Lee, Jinkwan Park, Daegeon Kwon, Bokuk Park, Hwan-Gue Cho, DoHoon Lee · 2015
A time series is a sequence of data sequentially observed over time. It contains important information such as stock, futures, foreign exchange rate, brainwave, heart rate, weather change, wind speed, air temperature and advective flux. Analyzing time series data is essential to estimate the future and understand scientific characteristics inherent in many phenomenon. There are a large number of methods for analyzing individual time series, especially on similarity measures. However, there is a lack in considering two time series simultaneously and discovering their correlation. In this paper, we propose a pairwise time-series walk plot model, which analyzes the correlation between two time series and visualizes both of them at the same time. And then we present a measure that represents the correlation of two time series by quantifying their complicated boundary shape.