Teams link https://teams.microsoft.com/meet/362259683825061?p=ufuPjWywp4jvY8VNN3
Add to Google Calendar
Abstract
This talk presents an overview of Multivariate Time Series Anomaly Detection (MTSAD) and recent developments in the literature, examining the main approaches and trends in the field. It also discusses current challenges in benchmarking and evaluation, exploring whether reported performance improvements truly reflect methodological progress or are influenced by inconsistencies in evaluation practices.
Short Bio
Sónia Gouveia is a researcher at the University of Aveiro, with a strong focus on developments in Probability and Statistics. Her work lies at the intersection of theory and real-world applications, combining methodological development with practical applications across many data settings. She places particular emphasis on developing statistical models and machine learning techniques applied to time series data.
She conducts her research at the R&D Unit IEETA in Computer Science, where she coordinates the Information Systems and Processing (ISP) research group. In parallel, she has extensive teaching experience across BSc, MSc and PhD levels, focusing on advanced training in statistical data analysis and modelling.
More information at News U. Porto