Rendezvous of Time Series Model and Temporal Data Mining

• Activity Number: 489
• Type: Contributed
• Date/Time: Wednesday, August 12, 2015 : 8:30 AM to 10:20 AM
• Sponsor: Business and Economic Statistics Section
• Author(s): Seong-Tae Kim*
• Keywords: Dynamic Time Warping ; temporal data mining ; similarity ; cointegration

Abstract

A number of methods in time series analysis focus on performing statistical hypothesis testing to uncover interdependent relationships among multiple time series processes. These methods are faced with some statistical issues such as misspecification, low testing power and curse of dimensionality. The purpose of this presentation is to examine the potential usage of a temporal data mining method, i.e., Dynamic Time Warping (DTW) algorithm to identify temporal interdependent relationships. DTW algorithms search optimal paths of various time series which determine similarity patterns. Analytical results show that DTW can capture various similarity patterns including cointegrated series, which could be a complementary tool of traditional time series methods.