SMS scnews item created by Linh Nghiem at Thu 24 Aug 2023 0856
Type: Seminar
Modified: Tue 29 Aug 2023 2135
Distribution: World
Expiry: 7 Sep 2023
Calendar1: 30 Aug 2023 1530-1630
CalLoc1: Carslaw 157
CalTitle1: Statistics seminar: Unsupervised Spatial-Temporal Decomposition for Feature Extraction and Anomaly Detection
Auth: linhn@110-174-229-71.tpgi.com.au (hngh7483) in SMS-SAML

Statistics Seminar

Statistics seminar: Unsupervised Spatial-Temporal Decomposition for Feature Extraction and Anomaly Detection

Liu

Next week, we will have two statistics seminars, one on Wednesday, August 30 and the other on Friday September 1. Both seminars are in the room Carslaw 157-257. This is the announcement for the first seminar on Wednesday August 30.

Title: Unsupervised Spatial-Temporal Decomposition for Feature Extraction and Anomaly Detection
Speaker: A/Prof Jian Liu, University of Arizona
Time and location: 3h30 PM - 4h30 PM Wednesday August 30 at Carslaw 157--257

Abstract: The advancement of sensing and information technology has made it reliable and affordable to collect data continuously from many sensors that are spatially distributed, generating readily available Spatial-Temporal (ST) data. While the abundant ST information embedded in such high-dimensional ST data provides engineers with unprecedented opportunities to understand, monitor, and control the engineering processes, the complex ST correlation makes conventional statistical data analysis methods ineffective and inefficient. This is especially true for ST feature extraction and ST anomaly detection, where the features of interest or the anomalies possess ST characteristics subtly different from the normal routine background. In this seminar, I will introduce a generic unsupervised learning method based on ST decomposition. The high-dimensional ST data are modeled as a tensor, which is then decomposed into different tensor components represented as a combination of a series of lower-dimensional factors. Without relying on pre-annotated training data, these tensor components will be estimated to indicate the latent features and/or anomalies of interest. A regularization approach is adopted to incorporate the knowledge of the tensor components’ intrinsic ST characteristics into the algorithm to improve the accuracy and robustness of the model estimation. Multiple case studies were conducted to demonstrate the effectiveness of the proposed methods, including water burst detection in water distribution systems and video segmentation.

Bio: Dr. Jian Liu is an Associate Professor in the Department of Systems & Industrial Engineering at The University of Arizona. Dr. Liu’s research specialty is the fusion of multi-source, multi-scale, and multilevel information in hierarchical and distributed systems for better system design, operation, and maintenance. He is a member of the Institute for Operations Research and Management Science (INFORMS) and a member of the Institute of Industrial and Systems Engineering (IISE). He served as a Council Member of the Quality, Statistics, and Reliability (QSR) Section of INFORMS from 2012 to 2014, the Chair of QSR from 2022 to 2023, a Board Director of the Quality Control and Reliability Engineering (QCRE) Division of IISE from 2013 to 2015, and the President of QCRE from 2016 to 2017. He serves as an Associate Editor of the Journal of Manufacturing Systems and an Associate Editor of the IISE Transactions. His research has been supported by NSF, and AFOSR, among others.