Machine Learning for Signal Processing 2013


By the local organising team for MLSP 2013


We had the pleasure to host the IEEE International Workshop on Machine Learning for Signal Processing (MLSP) at a retreat hotel in Chiltern Manor in Southampton. This event was special for two reasons. Firstly, MLSP returned to the United Kingdom after more than 15 years. Secondly, it was the first time that the MLSP community and the Independent Component Analysis/Latent Variable Analysis (ICA/LVA) community worked together to host a joint workshop. Researchers from all over the world contributed to this workshop, such as those from Princeton University, Stanford University, and Cambridge University.

The organisation of this workshop was a bit of a challenge, given that it was hosted away from University of Surrey. Thankfully, our dedicated team, including Saeid Sanei, Konstantinos Eftaxias, Samaneh Kouchaki, Shahrzad Shapoori, Shirin Enshaeifar and Clive Cheong Took, worked in collaboration with Chiltern Manor Hotel, particularly with Karen Welsh to make MLSP 2013 a success. The workshop is atypical from other workshops given that it favours “retreat” places such as Chiltern Manor in Southampton or Kittila in Finland away from the buzz of noisy cities. It is also a tradition for the conference to host an annual online competition on machine learning, which made the workshop fun! 79 data analysts participated in the competition to predict a set of bird species from a given data set. The recipient of the first prize was  Gabor Tobar from Hungary. He used random forests to perform the classification and won a prize of made of $700 and free registration. The second prize of $600 was won by Hong Wei Ng and Thi Ngoc Tho Nguyen.

The workshop kick-started with two tutorials on tensor tools for signal processing and on methods and applications on kernel Hilbert spaces. It was a nice surprise to see that the number of participants for those two tutorials was well above 50 delegates on the first day of the workshop. On the following days, eminent professors including Prof. S. Godsill, Prof. S. Theodoridis and Prof. C. Jutten delivered keynote talks on hot topics in machine learning for signal processing, such as sequential inference, sparsity aware learning algorithms, and source separation. The banquet was hosted on a cruise by Princess Carolina ship based in Southampton’s Ocean Village. Needless to say that the delegates looked forward to this event, after three days of lectures/poster presentations and fruitful discussions. We were happy that many delegates who were not planning to come for the banquet decided to join us for this well-deserved break!

Last but not least, we would like to thank the IEEE technical committee of Machine Learning for Signal Processing for giving us the opportunity and technical support to host this workshop. Also, a big thank you to the Department of Computing and other departments within University of Surrey such as the Finance Office for providing the administrative support, which made the workshop a success.