SDM 2015 Tutorial
Finding Repeated Structure in Time Series: Algorithms and Applications
Abdullah Mueen and Eamonn Keogh
Presentation slides are available here
Repeated patterns in time series data are indicative to identical dynamics in the origin. Such patterns can be used to summarize, classify, compress, cluster and classify time series data. In this tutorial, we will present several algorithms for repeated pattern discovery in univariate and multivariate time series data. The algorithms cover a wide range of settings from in-memory to online data, from approximate to exact algorithms and, from one length to all lengths. We will present applications of repeated patterns in several domains and in various data types. We will cover
* Exact and approximate algorithms for finding repeated patterns in time series
* Clustering, classification and rule discovery algorithms using repeated patterns
* Applications to Entomology, Data center management, Activity recognition
The tutorial will conclude with a list of open problems and research directions.
Dr. Mueen is an Assistant Professor at Computer Science in University of New Mexico. Dr. Mueen's research interest is in large-scale heterogeneous data mining with special focus on time series data. He is the runner-up of SIGKDD doctoral dissertation award in 2012. Dr. Mueen has won the best paper award in SIGKDD conference in 2012. Dr. Mueen is one of the authors of the most cited paper of SDM 2009 till now.
Dr. Keogh is a prolific author in data mining conferences, as of June 2014, he has 18 papers in SDM, making him one of the most prolific SDM authors. He has won best paper awards at ICDM, SIGKDD and SIGMOD. While he is only 13 years out from his PhD he has already obtained an H-index of 57. He has given well-received tutorials at SIGKDD (three times), ICDM (three time), VLDB, SDM, ACM Multimedia and CIKM.