Harri Valpola

Academy Research Fellow (2005-2010), Dr.Tech.

Postal Address:
Aalto University
School of Science and Technology
Dep. of Biomedical Engineering and Computatonal Science
P.O. Box 12200
FI-00076 AALTO, FINLAND
Visiting Address:
Rakentajanaukio 2, Espoo
3rd floor, Room F303
Tel:
+358 9 470 25724
Fax:
+358 9 470 23182
Email:
Harri.Valpola [at] tkk.fi

I have moved to ZenRobotics Ltd.. I'm still working part time at Aalto University, but I went back to ICS.

Between November 2004 and November 2010, I run the Computational Neuroscience group in the LCE at TKK. Before coming to the LCE, I worked for 14 months for the AI Lab at the University of Zurich and for more than ten years at the Neural Networks Research Centre at TKK.

Research Activities

In NNRC I was working on many methods for unsupervised learning. In order to widen my scope, I went to AI Lab to work with robots. I'm interested in natural and artificial intelligence and my current research is about "building a brain" (not the brain) for a robot. By brain I mean here an integrated control architecture which is capable of autonomous behaviour and learning. The motivation for this approach is that high-level cognitive control uses many of the same mechanisms as motor control. It is easier to study and understand navigation, manipulation and very basic (emotional) communication than planning, reasoning and symbolic language, but the same mechanisms are at work: we grasp objects and ideas, we find places and solutions. I believe that to get a complete picture of intelligence, one has to consider the whole behaviour and the interactions between individual adaptive systems that create it.

My speciality is learning representations (abstractions, concepts, etc.). Suitable internal representations are the essential ingredient for the transition from immediate motor control to high-level cognition. At the moment, one of my main interests is task-adapted sensory processing which includes development of behaviourally meaningful representations and attention (see self-organization of invariant representations and attention).

Of course it is possible to study task-adapted sensory processing only if there is a task. Since the brain has evolved to control movements, it is easiest to understand the brain by considering the challenges posed by autonomous robots and motor control. I've been studying mainly predictive motor control and reinforcement learning.

Here is more information about my research.

Publications

Below is a list of my recent publications. Click the titles for descriptions of the publications and related links. The topic of most of my publications is (broadly speaking) either cognitive architecture, denoising source separation or variational Bayesian learning.

2010
A cognitive architecture for developing sensory and motor abstractions
H. Valpola
A presentation given at the First International Conference on Biologically Inspired Cognitive Architectures, BICA 2010.
Oscillatory neural network for image segmentation with biased competition for attention
T. Raiko and H. Valpola.
In the Brain Inspired Cognitive Systems (BICS 2010) symposium, Madrid, Spain, 14-16 July, 2010.
2009
Selective attention improves learning
A. Yli-Krekola, J. Särelä and H. Valpola.
In Proceedings of the 19th International Conference of Artificial Neural Networks, ICANN 2009, Limassol, Cyprus, Part II, pp. 285-294, 2009.
2008
From raw data to abstract concepts
H. Valpola
Keynote presentation in AKRR'08: International and Interdisciplinary Conference on Adaptive Knowledge Representation and Reasoning, Porvoo, Finland, September 17-19, 2008.
The engine of thought -- a bio-insipred mechanism for distributed selection of useful information
H. Valpola.
Nokia Workshop on Machine Consciousness, Helsinki, Finland, pp. 27-31, 2008.
2007
Computational model of co-operating covert attention and learning.
A. Yli-Krekola and H. Valpola.
Fifth Nordic Neuroinformatics Workshop, Espoo, Finland, p. 34, 2007.
A model of cerebellar automation of voluntary basal-ganglia control.
M. Pihlaja and H. Valpola.
Fifth Nordic Neuroinformatics Workshop, Espoo, Finland, p. 29, 2007.
Cerebellar model tested in control of a load-carrying robot.
I. Aaltonen and H. Valpola.
Fifth Nordic Neuroinformatics Workshop, Espoo, Finland, p. 16, 2007.
Cerebellar model for coordination.
T. J. Lukka and H. Valpola.
Fifth Nordic Neuroinformatics Workshop, Espoo, Finland, p. 25, 2007.
Compact modeling of data using independent variable group analysis.
E. Alhoniemi, A. Honkela, K. Lagus, J. Seppä, P. Wagner and H. Valpola.
IEEE Transactions on Neural Networks, 18(6):1762-1776, 2007.
Blind separation of nonlinear mixtures by variational Bayesian learning.
A. Honkela, H. Valpola, A. Ilin and J. Karhunen.
Digital Signal Processing, 17(5):914-934, 2007.
Building blocks for variational Bayesian learning of latent variable models.
T. Raiko, H. Valpola, M. Harva and J. Karhunen.
Journal of Machine Learning Research, 8:155-201, 2007.
Finding interesting climate phenomena by exploratory statistical techniques.
A. Ilin, H. Valpola and E. Oja.
In Proceedings of the Fifth Conference on Artificial Intelligence Applications to Environmental Science, 5AI, as part of the 87th Annual Meeting of the American Meteorological Society, San Antonio, TX, USA, January 2007.
2006
Hyperparameter adaptation in variational Bayes for the gamma distribution.
H. Valpola and A. Honkela.
Helsinki University of Technology, Publications in Computer and Information Science, Espoo, Finland, Tech. Rep. E6, 2006.
Learning anticipatory behaviour using a simple cerebellar model.
H. Valpola.
In Proceedings of the Ninth Scandinavian Conference on Artificial Intelligence, SCAI 2006, Espoo, Finland, pp. 135-142, 2006.
Extraction of climate components with structured variance.
A. Ilin, H. Valpola and E. Oja.
In Proceedings of the IEEE World Congress on Computational Intelligence, WCCI 2006, Vancouver, BC, Canada, pp. 10528-10535, 2006.
Exploratory analysis of climate data using source separation methods.
A. Ilin, H. Valpola and E. Oja.
Neural Networks, 19(2):155-167, 2006.
Separation of nonlinear image mixtures by denoising source separation.
M. S. C. Almeida, H. Valpola and J. Särelä.
In Proceedings of the 6th International Conference on Independent Component Analysis and Blind Signal Separation, ICA 2006, Charleston, SC, USA, pp. 8-15, 2006.
2005
Frequency-based separation of climate signals.
A. Ilin and H. Valpola.
In Proceedings of the 9th European Conference on Principles and Practice of Knowledge Discovery in Databases (PKDD 2005), Porto, Portugal, pp. 519-526, 2005.
On the effect of the form of the posterior approximation in variational learning of ICA models.
A. Ilin and H. Valpola.
Neural Processing Letters 22(2):183-204, 2005.
Semiblind source separation of climate data detects El Niño as the component with the highest interannual variability.
A. Ilin, H. Valpola and E. Oja.
In Proceedings of the International Joint Conference on Neural Networks (IJCNN 2005), Montréal, Québec, Canada, pp. 1722-1727, 2005.
Bayes Blocks: An implementation of the variational Bayesian building blocks framework.
M. Harva, T. Raiko, A. Honkela, H. Valpola and J. Karhunen.
In Proceedings of the 21st Conference on Uncertainty in Artificial Intelligence (UAI 2005), Edinburgh, Scotland, pp. 259-266, 2005.
Development of representations, categories and concepts--a hypothesis.
H. Valpola.
In Proceedings of the 6th IEEE International Symposium on Computational Intelligence in Robotics and Automation, CIRA 2005, Espoo, Finland, pp. 593-599, 2005.
Unsupervised variational Bayesian learning of nonlinear models.
A. Honkela and H. Valpola.
In L. Saul, Y. Weis and L. Bottous, eds., Advances in Neural Information Processing Systems 17 (NIPS 2004), pp. 593-600, 2005.
Denoising source separation: a novel approach to ICA and feature extraction using denoising and Hebbian learning.
J. Särelä and H. Valpola.
In AI 2005 special session on correlation learning, pp. 45-56, 2005.
Denoising source separation.
J. Särelä and H. Valpola.
Journal of Machine Learning Research 6:233-272, 2005.
2004
Accurate, fast and stable denoising source separation algorithms.
H. Valpola and J. Särelä.
In Proceedings of the 5th International Conference on Independent Component Analysis and Blind Signal Separation (ICA 2004), Granada, Spain, pp. 65-72, 2004.
Using kernel PCA for initialisation of variational Bayesian nonlinear blind source separation method.
A. Honkela, S. Harmeling, L. Lundqvist and H. Valpola
In Proceedings of the 5th International Conference on Independent Component Analysis and Blind Signal Separation (ICA 2004), Granada, Spain, pp. 65-72, 2004.
Denoising source separation: from temporal to contextual invariance.
H. Valpola and J. Särelä.
Presented in Early Cognitive Vision Workshop, Isle of Skye, Scotland, 2004.
Behaviourally meaningful representations from normalisation and context-guided denoising.
H. Valpola.
AI Lab technical report, University of Zurich, 2004.
Variational learning and bits-back coding: an information theoretic view to Bayesian learning.
A. Honkela and H. Valpola.
IEEE Transactions on Neural Networks, 15(4):800-810, 2004.
Nonlinear dynamical factor analysis for state change detection.
A. Ilin, H. Valpola and E. Oja.
IEEE Transactions on Neural Networks, 15(3):559-575, 2004.
Hierarchical models of variance sources.
H. Valpola, M. Harva and J. Karhunen.
Signal Processing, 84(2):267-282, 2004.
1995-2003

Most of my publications up to 2003 are available only at my NNRC page.

An unsupervised ensemble learning method for nonlinear dynamic state-space models.
H. Valpola and J. Karhunen.
Neural Computation, 14(11):2647-2692, 2002.
A fast semi-blind source separation algorithm.
H. Valpola and J. Särelä.
In Publications in Computer and Information Science, Report A66, Helsinki University of Technology, Espoo, Finland, 4 p., 2002.
Bayesian ensemble learning for nonlinear factor analysis.
H. Valpola.
PhD thesis, Helsinki University of Technology, Espoo, 2000.
Published in Acta Polytechnica Scandinavica, Mathematics and Computing Series No. 108, 2000.
Fast algorithms for Bayesian independent component analysis.
H. Valpola and P. Pajunen.
In Proceedings of the Second International Workshop on Independent Component Analysis and Blind Signal Separation, ICA 2000, Helsinki, Finland, pp. 233-237, 2000.

Teaching

I'm now lecturing the course Tfy-99.3730 Information processing in the brain.

In spring 2008, I gave two lectures for the course Tfy-99.4247 Ihmisaivojen rakenne ja toiminta (in Finnish)

In autumn 2006, I lectured the course S-114.3812 Computational Neuroscience in collaboration with Jarmo Hurri from Helsinki University.

In spring 2006, I lectured the following course: S-114.4220 Research Seminar on Computational Science. The topic was: Principles of brain evolution.

In spring 2005, I lectured the following course: S-114.220 Research Seminar on Computational Science. The topic was neurorobotics.

I'm also involved in Synthetic Brain which is a project started by people who were involved in NeuroHel, a multidisciplinary neuroscience study group.

Other Info


Harri Valpola
Last modified: Tue Nov 10 15:15:51 EET 2009