MotifTICC: Motif-Aware Toeplitz Inverse Covariance-Based Clustering in Time Series Data

<a href="https://embed.stanford.edu/iframe/?url=https%3A%2F%2Fpurl.stanford.edu%2Fwz520xh6722" class="su-underline">Show Content</a>

Abstract/Contents

Abstract: Clustering and segmentation in multivariate time-series datasets is useful for understanding a system's behavior over time. In particular, the task of finding repeated heterogeneous patterns, or motifs, is helpful for interpreting large datasets and identifying interesting patterns. Performing hierarchical clustering is challenging because it involves simultaneously learning both a primary clustering layer and an overlying motif set. In this paper, we propose MotifTICC, an algorithm for discovering noisy motifs and hierarchically clustering time series data. MotifTICC both increases the robustness of the primary clustering layer and outputs a ranked list of motifs. We validate this algorithm on synthetic data and apply the method to an automobile dataset.

Type of resource	text
Date created	[ca. June 5, 2018]

Author	Jain, Saachi
Advisor	Leskovec, Jure
Degree granting institution	Stanford University, Department of Computer Science

Location	https://purl.stanford.edu/wz520xh6722

Use and reproduction: User agrees that, where applicable, content will not be used to identify or to otherwise infringe the privacy or confidentiality rights of individuals. Content distributed via the Stanford Digital Repository may be subject to additional license and use restrictions applied by the depositor.
License: This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).

Preferred Citation: Jain, Saachi. (2018). MotifTICC: Motif-Aware Toeplitz Inverse Covariance-Based Clustering in Time Series Data. Stanford Digital Repository. Available at: https://purl.stanford.edu/wz520xh6722

Undergraduate Theses, School of Engineering

Loading usage metrics...