Stacked autoencoders-based representation learning for temporal and trajectory data

• Grant number: 19/06080-2
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date until: July 01, 2019
• End date until: August 31, 2020
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Diego Furtado Silva
• Grantee: Yuri Gabriel Aragão da Silva
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil

Abstract

With the development of mobile sensors, applications that collect large volumes of temporal data have emerged in different knowledge domains. In many of these applications, using algorithms to find frequent and anomalous patterns is an alternative to discover knowledge from data. In the case of high-dimensional temporal data, the literature indicates that similarity measures more suitable for several applications have high computational complexity. Two options in this context are to use less expensive distance measures or approximations, which may negatively affect the results obtained. As a way to overcome this problem, this project proposes using autoencoders to learn reduced representations of time series to improve the scalability of similarity-based pattern discovery algorithms without causing significant loss of quality.

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

ARAGAO DA SILVA, YURI GABRIEL; SILVA, DIEGO FURTADO; WU, XT; JERMAINE, C; XIONG, L; HU, XH; KOTEVSKA, O; LU, SY; XU, WJ; ALURU, S; et al. On Convolutional Autoencoders to Speed Up Similarity-Based Time Series Mining. 2020 IEEE INTERNATIONAL CONFERENCE ON BIG DATA (BIG DATA), v. N/A, p. 10-pg., 2020-01-01. (19/06080-2, 17/24340-6)