Spatiotemporal Time-Series Analysis Using Kernel-Based Methods

Ben Naftali Hen, M.Sc. student at the department of Industrial Engineering 
Advisor: Dr. Neta Rabin 

Via Zoom click here

Abstract:
Spatio-temporal time-series analysis is a growing area of research that includes different types of tasks, such as forecasting, prediction, clustering, and visualization. In many domains, like epidemiology or economics, time series data is collected in order to describe the observed phenomenon in particular locations over a predefined time slot and predict future behavior. Regression methods provide a simple mechanism for evaluating empirical functions over scattered data points. In particular, kernel-based regressions are suitable for cases in which the relationship between the data points and the function is not linear. In this work, we propose a kernel-based iterative regression model, based on the Auto-Adaptive Laplacian Pyramids (ALP), which fuses data from several spatial locations for improving the forecasting accuracy of a given time series. In more detail, the proposed method approximates and extends a function based on two or more spatial input modalities coded by a series of multiscale kernels, which are averaged as a convex combination. In addition, we propose the EMD-ALP, a hybrid approach that combines ALP with Empirical Mode Decomposition (EMD), for constructing more accurate ALP models. Experimental results demonstrate the proposed models for solar energy prediction, forecasting epidemiology infections and future number of fire events. Both methods are compared with wellknown regression techniques and highlights the benefits of the proposed kernel-based regression in terms of accuracy and flexibility.

Bio:

Ben Hen is an M.Sc student at the department of Industrial Engineering at Tel Aviv University, specializing in Data Science. Ben holds a B.Sc degree in Industrial Engineering and Management also from Tel Aviv University. Ben has experience as a Fraud Data Analyst at Global-E. This work was conducted under the supervision of Dr. Neta Rabin.

E-Mail: Benhen@mail.tau.ac.il

Linkedin: https://www.linkedin.com/in/benhen23