From nicolas.brodu at numerimoire.net Tue Oct 6 19:03:33 2009 From: nicolas.brodu at numerimoire.net (Nicolas Brodu) Date: Tue, 6 Oct 2009 21:03:33 +0200 Subject: [Causality-ML] Reminder: Telling cause from effect based on high-dimensional observations Message-ID: <200910062103.33382.nicolas.brodu@numerimoire.net> Dear Causality and Machine Learning group, This is a friendly reminder for the next presentation is by Dominik Janzing: Title: Telling cause from effect based on high-dimensional observations When: Thursday 8 October, 17h Paris time (8h PT, 11h ET, 23h Beijing) URL: http://www.afia-france.org/tiki-index.php?page=GroupeDeLecture091008DJ Participation is as simple as opening a URL in a recent web browser (ex: Firefox), and optionally calling a local phone number. See the usage help at http://www.afia-france.org/tiki-index.php?page=teleconfConnectionDetails You may also like to check our sister project on complex systems at http://teleconf.csregistry.org Abstract of the presentation: -------- We describe a method for inferring linear causal relations among multi-dimensional variables. The idea is to use an asymmetry between the distributions of cause and effect that occurs if both the covariance matrix of the cause and the structure matrix mapping cause to the effect are independently chosen. The method works for both stochastic and deterministic causal relations, provided that the dimensionality is sufficiently high (in some experiments, 5 was enough). It is applicable to Gaussian as well as non-Gaussian data. -------- If you know of potentially interested speakers or if you wish to present a paper, please send us a message so we can add you in the planning. The schedule for the next presentations is maintained at: http://www.afia-france.org/tiki-index.php?page=Groupe+de+lecture Best regards, Nicolas Brodu From nicolas.brodu at numerimoire.net Mon Oct 12 21:56:59 2009 From: nicolas.brodu at numerimoire.net (Nicolas Brodu) Date: Mon, 12 Oct 2009 23:56:59 +0200 Subject: [Causality-ML] Replay and next presentation: Approximating viability kernels with support vector machines Message-ID: <200910122356.59863.nicolas.brodu@numerimoire.net> Dear Causality and Machine Learning group, The video replay of the last presentation is available at: http://www.encours.org/causality/causalityML091008DJ/replay.html The next presentation is by Guillaume Deffuant: Title: Approximating viability kernels with support vector machines When: Thursday 15 October, 17h Paris time (8h PT, 11h ET, 23h Beijing) Reference: IEEE Trans. on Automatic Control (52) 5, pp 933-937. URL: http://www.encours.org/encours/?presentation=causalityML091015GD Participation is as simple as opening the above URL in a recent web browser (ex: Firefox), and optionally calling a local phone number. See the usage help at http://www.afia-france.org/tiki-index.php?page=teleconfConnectionDetails You may also like to check our sister project on complex systems at http://teleconf.csregistry.org Abstract of the presentation: -------- We propose an algorithm which performs a progressive approximation of a viability kernel, iteratively using a classification method. We establish the mathematical conditions that the classification method should fulfil to guarantee the convergence to the actual viability kernel. We study more particularly the use of support vector machines (SVMs) as classification techniques. We show that they make possible to use gradient optimisation techniques to find a viable control at each time step, and over several time steps. This allows us to avoid the exponential growth of the computing time with the dimension of the control space. It also provides simple and efficient control procedures. We illustrate the method with some examples inspired from ecology. -------- If you know of potentially interested speakers or if you wish to present a paper, please send us a message so we can add you in the planning. The schedule for the next presentations is maintained at: http://www.afia-france.org/tiki-index.php?page=Groupe+de+lecture Best regards, Nicolas Brodu From nicolas.brodu at numerimoire.net Thu Oct 15 11:38:22 2009 From: nicolas.brodu at numerimoire.net (Nicolas Brodu) Date: Thu, 15 Oct 2009 13:38:22 +0200 Subject: [Causality-ML] Reminder: Approximating viability kernels with support vector machines Message-ID: <200910151338.23334.nicolas.brodu@numerimoire.net> Dear Causality and Machine Learning group, This is a friendly reminder for today's presentation by Guillaume Deffuant: Title: Approximating viability kernels with support vector machines When: Thursday 15 October, 17h Paris time (8h PT, 11h ET, 23h Beijing) Reference: IEEE Trans. on Automatic Control (52) 5, pp 933-937. URL: http://www.encours.org/encours/?presentation=causalityML091015GD Participation is as simple as opening the above URL in a recent web browser (ex: Firefox), and optionally calling a local phone number. See the usage help at http://www.afia-france.org/tiki-index.php?page=teleconfConnectionDetails You may also like to check our sister project on complex systems at http://teleconf.csregistry.org Abstract of the presentation: -------- We propose an algorithm which performs a progressive approximation of a viability kernel, iteratively using a classification method. We establish the mathematical conditions that the classification method should fulfil to guarantee the convergence to the actual viability kernel. We study more particularly the use of support vector machines (SVMs) as classification techniques. We show that they make possible to use gradient optimisation techniques to find a viable control at each time step, and over several time steps. This allows us to avoid the exponential growth of the computing time with the dimension of the control space. It also provides simple and efficient control procedures. We illustrate the method with some examples inspired from ecology. -------- If you know of potentially interested speakers or if you wish to present a paper, please send us a message so we can add you in the planning. The schedule for the next presentations is maintained at: http://www.afia-france.org/tiki-index.php?page=Groupe+de+lecture Best regards, Nicolas Brodu From nicolas.brodu at numerimoire.net Sun Oct 18 22:22:25 2009 From: nicolas.brodu at numerimoire.net (Nicolas Brodu) Date: Mon, 19 Oct 2009 00:22:25 +0200 Subject: [Causality-ML] Analyzing Coherent Brain Networks with Granger Causality Message-ID: <200910190022.25376.nicolas.brodu@numerimoire.net> Dear Causality and Machine Learning group, The next presentation is by Mingzhou Ding: Title: Analyzing Coherent Brain Networks with Granger Causality When: Thursday 22 October, 17h Paris time (8h PT, 11h ET, 23h Beijing) URL: http://www.afia-france.org/tiki-index.php?page=GroupeDeLecture091022MD Participation is as simple as opening a URL in a recent web browser (ex: Firefox), and optionally calling a local phone number. See the usage help at http://www.afia-france.org/tiki-index.php?page=teleconfConnectionDetails You may also like to check our sister project on complex systems at http://teleconf.csregistry.org Abstract of the presentation: -------- Multielectrode neurophysiological recording and functional brain imaging produce massive quantities of data. Multivariate time series analysis provides the basic framework for analyzing the patterns of neural interactions in these data. It has long been recognized that neural interactions are directional. Being able to assess the directionality of neuronal interactions is thus a highly desired capability for understanding the cooperative nature of neural computation. Research over the last few years has shown that Granger causality is a key technique to furnish this capability. In this talk, I will introduce the concept of Granger causality and present results from applications of this technique to multichannel LFP and EEG recordings from monkeys and humans performing cognitive tasks. -------- If you know of potentially interested speakers or if you wish to present a paper, please send us a message so we can add you in the planning. The schedule for the next presentations is maintained at: http://www.afia-france.org/tiki-index.php?page=Groupe+de+lecture Best regards, Nicolas Brodu