Advanced Machine Learning Techniques for Pattern Recognition in Time Series (PRISMA)

Funded by: 
Ministerio de Economía y Competitividad (TEC2014-57402-JIN)
Ministerio de Economía y Competitividad

The goal of PRISMA is to advance the state of the art in machine learning theory and algorithms that exploit temporal information. PRISMA aims to develop a general probabilistic framework to deal with temporal dynamics in time series, based on Bayesian graphical models for temporal pattern recognition. Furthermore, PRISMA aims to extend the state-of-the-art in kernel methods and machine learning algorithms that deal explicitly with the temporal variable in pattern recognition problems. The developed theoretical framework will be evaluated in applications from several key areas of the modern digital society.

Publications for this project

Conference Paper


S. Van Vaerenbergh, S. Scardapane, and I. Santamaría, "Recursive multikernel filters exploiting nonlinear temporal structure", European Signal Processing Conference (EUSIPCO), Kos, Greece, August, 2017. PDF icon PDF Version (290.2 KB)


S. Van Vaerenbergh, D. Comminiello, and L. A. Azpicueta-Ruiz, "A Split Kernel Adaptive Filtering Architecture for Nonlinear Acoustic Echo Cancellation", 24th European Signal Processing Conference (EUSIPCO 2016), Budapest, Hungary, September, 2016. PDF icon PDF Version (374.7 KB)
S. Van Vaerenbergh, J. Fernández-Bes, and V. Elvira, "On The Relationship Between Online Gaussian Process Regression And Kernel Least Mean Squares Algorithms", 2016 IEEE International Workshop on Machine Learning for Signal Processing (MLSP 2016), Salerno, Italy, IEEE, September, 2016. PDF icon PDF Version (175.24 KB)