Christoph considered blind source separation methods for spatial data and Una non-Gaussian feature extraction for vector- and matrix-variate data. Both dissertations will be soon online available.
during the spring we published several new R packages on CRAN. KernelICA gives an implementation of kernel ICA, and ssaBSS has several methods for (non-)stationary source separation. shapeNA allows for robust shape estimation when there are missing values in the data.
Meanwhile I am involved in 22 packages on CRAN. Below a figure about their monthly download stats as provided from the package dlstats.
Since beginning of the month I am tenured!
At that occassion I would also like to point out that many new papers are out. See my publication list for details. To give an impression of my research I also added below a word cloud of the abstracts of my published articles.
I’m heading to ICORS as an invited speaker and will give a talk about robust approaches for blind source separation. This will be the first time I get to visit Ecuador!
Our paper “Determining the signal dimension in second order source separation” appeared now as online first in Statistica Sinica. You can read it here.
Great news! The Austrian Science Fund (FWF) will fund my
project “Blind Source Separation in Time and Space” with almost 400000 Euros.
It has been a while since my last post – several papers have appeared since then, like
- “Supervised Dimension Reduction for Multivariate Time Series” appeared in Econometrics and Statistics.
- “Independent Component Analysis for Tensor-valued Data” appeared in the Journal in the Multivariate Analysis. For a short time it can be downloaded at here.
- I presented our paper “Blind Source Separation For Nonstationary Tensor-Valued Time Series” at the IEEE International Workshop on Machine Learning for Signal Processing. It should appear soon IEEE Xplore.
More details are also on my list of publications.
Beginning of this month I started a tenure track position in computational statistics at the Institute of Statistics & Mathematical Methods in Economics at the Vienna University of Technology. I’m there a member of the CSTAT group of Peter Filzmoser. It is all quite exciting!
This is also the reason why there were recently no updates on this page – but some papers have been appeared meanwhile – for details see my publication list!
“Blind Source Separation Based on Joint Diagonalization in R: The Packages JADE and BSSasymp” has appeared
Our paper “Blind Source Separation Based on Joint Diagonalization in R: The Packages JADE and BSSasymp” has appeared in the Journal of Statistical Software!
Along with it come new version of our R packages JADE and BSSasymp which are both on CRAN.