From nicolas.brodu at numerimoire.net Wed Jun 3 21:28:50 2009 From: nicolas.brodu at numerimoire.net (Nicolas Brodu) Date: Wed, 3 Jun 2009 23:28:50 +0200 Subject: [Causality-ML] Tomorrow's talk: A Strategy for Making Predictions Under Manipulation Message-ID: <200906032328.50963.nicolas.brodu@numerimoire.net> Dear Causality and Machine Learning group, Tomorrow's presentation by Laura Brown is about the strategy she and Ioannis Tsamardinos used for the LOCANET task from the Pot-luck challenge and NIPS 2008 workshop: When: Thursday 04 June, Paris 17h, ET 11h, PT 8h, Tokyo 0h(Friday) URL: http://www.afia-france.org/tiki-index.php?page=GroupeDeLecture090604LB Phone number: +1 (218) 936-7999 Participant code: 665140# See below for tips how to access the conference on the Internet using voice-over-IP. Abstract of the presentation: -------- The LOCANET challenge given to participants was to learn the local causal structure around a given target variable out to depth 3. We discuss the importance of this challenge for several research questions. The challenge used the unmanipulated data sets also used in the Causation and Prediction Challenge. The main technique used in this challenge was an approach to learning local regions that involves first identifying the variables in the region out to depth 4 (requested depth + 1). The local neighborhoods are then used to restrict a search for a final structure. The results of our method versus the reference methods and other participants will be presented. Finally, we intend to discuss lessons learned from the results of the challenge and how these lessons may lead to future directions for analysis and research avenues. -------- 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 planning for the next presentation is maintained at: http://www.afia-france.org/tiki-index.php?page=Groupe+de+lecture Tip for a cheap voice-over-IP connection to the conference: Open an account on nonoh.net (server in Germany) or voipcheap.com (server in US) or another SIP provider near your country. Pay 10 euros for 3 or 4 month unlimited access to nearly all land line phone numbers in the world, including the teleconference number. This means you will be able to spend as much time as you want for free on long-distance and international calls with the basic fee. Use a microphone + headset on your computer, and a program like Twinkle (Linux) or Ekiga (multi-OS) or X-Lite (windows). Set it up with your nonoh.net or voipcheap account (see the help page on these sites). Beware of filtering firewalls. Then call our teleconference number from your computer. Note: We are not affiliated to any of these groups. We provide this information in the hope it is useful. Best regards, Nicolas Brodu From nicolas.brodu at numerimoire.net Sun Jun 7 08:30:37 2009 From: nicolas.brodu at numerimoire.net (Nicolas Brodu) Date: Sun, 7 Jun 2009 10:30:37 +0200 Subject: [Causality-ML] Video replays for the presentations of Changzang Wang and Laura Brown Message-ID: <200906071030.38038.nicolas.brodu@numerimoire.net> Dear Causality and Machine Learning group, The following videos about the LOCANET task of the causality challenge are now available online and for download: - Changzang Wang: "Discover Local Causal Network around a Target with a Given Depth" http://www.encours.org/causality/locanet_pcd/replay.html - Laura Brown: "Learning Local Regions of Causal Networks" http://www.encours.org/causality/locanet_brown/replay.html The next presentation will be on Thursday 11th June by Ernest Mwebaze: "Fast Committee-Based Structure Learning". This will conclude our series on the LOCANET task. Thanks for your interest, Nicolas Brodu From nicolas.brodu at numerimoire.net Tue Jun 9 13:38:32 2009 From: nicolas.brodu at numerimoire.net (Nicolas Brodu) Date: Tue, 9 Jun 2009 15:38:32 +0200 Subject: [Causality-ML] Next talk: Fast Committee-Based Structure Learning Message-ID: <200906091538.32241.nicolas.brodu@numerimoire.net> Dear Causality and Machine Learning group, The next presentation will be on Thursday 11, by Ernest Mwebaze. This talk will conclude our series on the LOCANET task from the Pot-luck challenge and NIPS 2008 workshop: When: Thursday 11 June, Paris 17h, ET 11h, PT 8h, Tokyo 0h(Friday) URL: http://www.afia-france.org/tiki-index.php?page=GroupeDeLecture090611EM Phone number: +1 (218) 936-7999 Participant code: 665140# See below for tips how to access the conference on the Internet using voice-over-IP. Abstract of the presentation: -------- We present a Fast Committee-Based Structure learning method that was used in the LOCANET challenge. Current methods for causal structure learning tend to be computationally intensive or intractable for large datasets. Some recent approaches have speeded up the process by first making hard decisions about the set of parents and children for each variable, in order to break large-scale problems into sets of tractable local neighbourhoods. We use this principle in order to apply a structure learning committee for orientating edges between variables. We find that a combination of weak structure learners can be effective in recovering causal dependencies. Though such a formulation would be intractable for large problems at the global level, we show that it can run quickly when processing local neighbourhoods in turn. Experimental results show that this localized, committee-based approach has advantages over standard causal discovery algorithms both in terms of speed and accuracy. -------- 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 planning for the next presentation is maintained at: http://www.afia-france.org/tiki-index.php?page=Groupe+de+lecture Tip for a cheap voice-over-IP connection to the conference: Open an account on nonoh.net (server in Germany) or voipcheap.com (server in US) or another SIP provider near your country. Pay 10 euros for 3 or 4 month unlimited access to nearly all land line phone numbers in the world, including the teleconference number. This means you will be able to spend as much time as you want for free on long-distance and international calls with the basic fee. Use a microphone + headset on your computer, and a program like Twinkle (Linux) or Ekiga (multi-OS) or X-Lite (windows). Set it up with your nonoh.net or voipcheap account (see the help page on these sites). Beware of filtering firewalls. Then call our teleconference number from your computer. Note: We are not affiliated to any of these groups. We provide this information in the hope it is useful. Best regards, Nicolas Brodu From A.P.Dawid at statslab.cam.ac.uk Thu Jun 11 19:00:57 2009 From: A.P.Dawid at statslab.cam.ac.uk (Philip Dawid) Date: Thu, 11 Jun 2009 20:00:57 +0100 (BST) Subject: [Causality-ML] Revised paper Message-ID: Dear friends, At the suggestion of Isabelle Guyon, I am making available, as an attachment to this message, the revised version of my paper "Beware of the DAG!". Please note that this has been very considerably rewritten in the light of the many helpful comments received, both in writing and in the oral discussion of my web-seminar. I should be grateful if you would refrain from any further comments on the original version, though by all means feel free to lay into this one (though whether it will be the final version is currently in the hands of the referees...) I should like to take this opportunity to thank all those who have contributed suggestions and ideas, and helped me to clarify my arguments (to myself, at any rate). Philip ------------------------------------------------------------------------- A. P. Dawid, Professor of Statistics, Statistical Laboratory, Centre for Mathematical Sciences, Wilberforce Road, Cambridge CB3 0WB, UK Tel: +44 (0)1223 766535 E-mail: apd[at]statslab.cam.ac.uk Personal web page: http://tinyurl.com/2maycn Department web page: http://www.statslab.cam.ac.uk/ ------------------------------------------------------------------------- -------------- next part -------------- A non-text attachment was scrubbed... Name: apd_jmlr.pdf Type: application/pdf Size: 459402 bytes Desc: apd_jmlr.pdf URL: From A.P.Dawid at statslab.cam.ac.uk Thu Jun 11 19:32:38 2009 From: A.P.Dawid at statslab.cam.ac.uk (Philip Dawid) Date: Thu, 11 Jun 2009 20:32:38 +0100 (BST) Subject: [Causality-ML] Revised paper -- please replace with this file In-Reply-To: References: Message-ID: Uggghhh!!!! --- wrong file attached (that was the original version). Trying again... Apologies, Philip ------------------------------------------------------------------------- A. P. Dawid, Professor of Statistics, Statistical Laboratory, Centre for Mathematical Sciences, Wilberforce Road, Cambridge CB3 0WB, UK Tel: +44 (0)1223 766535 E-mail: apd[at]statslab.cam.ac.uk Personal web page: http://tinyurl.com/2maycn Department web page: http://www.statslab.cam.ac.uk/ ------------------------------------------------------------------------- On Thu, 11 Jun 2009, Philip Dawid wrote: > > Dear friends, > > At the suggestion of Isabelle Guyon, I am making available, as an attachment > to this message, the revised version of my paper "Beware of the DAG!". > Please note that this has been very considerably rewritten in the light of > the many helpful comments received, both in writing and in the oral > discussion of my web-seminar. I should be grateful if you would refrain from > any further comments on the original version, though by all means feel free > to lay into this one (though whether it will be the final version is > currently in the hands of the referees...) > > I should like to take this opportunity to thank all those who have > contributed suggestions and ideas, and helped me to clarify my > arguments (to myself, at any rate). > > Philip > > ------------------------------------------------------------------------- > A. P. Dawid, Professor of Statistics, Statistical Laboratory, > Centre for Mathematical Sciences, Wilberforce Road, Cambridge CB3 0WB, UK > Tel: +44 (0)1223 766535 > E-mail: apd[at]statslab.cam.ac.uk > Personal web page: http://tinyurl.com/2maycn > Department web page: http://www.statslab.cam.ac.uk/ > ------------------------------------------------------------------------- > -------------- next part -------------- A non-text attachment was scrubbed... Name: apd_jmlr_rev.pdf Type: application/pdf Size: 477146 bytes Desc: apd_jmlr_rev.pdf URL: From nicolas.brodu at numerimoire.net Mon Jun 15 23:50:37 2009 From: nicolas.brodu at numerimoire.net (Nicolas Brodu) Date: Tue, 16 Jun 2009 01:50:37 +0200 Subject: [Causality-ML] This week's talk: Inferring acyclic network structures using Granger causality tests Message-ID: <200906160150.37691.nicolas.brodu@numerimoire.net> Dear Causality and Machine Learning group, The next presentation will be on Thursday 18 (this week), by Radhakrishnan Nagarajan. When: Thursday 18 June, Paris 17h, ET 11h, PT 8h, Tokyo 0h(Friday) URL: http://www.afia-france.org/tiki-index.php?page=GroupeDeLecture090618RN Phone number: +1 (218) 936-7999 Participant code: 665140# Reference paper: ---------------- Nagarajan, R (2009): A Note on inferring acyclic network structures using Granger causality tests. The International Journal of Biostatistics. 5(1), Article 10. Abstract of the presentation: ----------------------------- Granger causality (GC) and its extension have been used widely to infer causal relationships from multivariate time series generated from biological systems. GC is ideally suited for causal inference in bivariate vector autoregressive process (VAR). A zero magnitude of the upper or lower off-diagonal element(s) in a bivariate VAR is indicative of lack of causal relationship in that direction resulting in true acyclic structures. However, in experimental settings, statistical tests, such as F-test that rely on the ratio of the mean-squared forecast errors, are used to infer significant GC relationships. The presentation will investigate acyclic approximations within the context of bi-directional two-gene network motifs modeled as bivariate VAR. The fine interplay between the model parameters in the bivariate VAR, namely: (i) transcriptional noise variance, (ii) autoregulatory feedback, and (iii) transcriptional coupling strength that can give rise to discrepancies in the ratio of the mean-squared forecast error are elucidated. Subsequently, their impact on statistical power is investigated using Monte Carlo simulations. While the results presented are generic, the presentation is likely to have significant impact on inferring functional relationships and transcriptional circuitry from gene expression profiles. ----------------------------- 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 planning for the next presentation is maintained at: http://www.afia-france.org/tiki-index.php?page=Groupe+de+lecture Tip for a cheap voice-over-IP connection to the conference: Open an account on nonoh.net (server in Germany) or voipcheap.com (server in US) or another SIP provider near your country. Pay 10 euros for 3 or 4 month unlimited access to nearly all land line phone numbers in the world, including the teleconference number. This means you will be able to spend as much time as you want for free on long-distance and international calls with the basic fee. Use a microphone + headset on your computer, and a program like Twinkle (Linux) or Ekiga (multi-OS) or X-Lite (windows). Set it up with your nonoh.net or voipcheap account (see the help page on these sites). Beware of filtering firewalls. Then call our teleconference number from your computer. Note: We are not affiliated to any of these groups. We provide this information in the hope it is useful. Best regards, Nicolas Brodu From nicolas.brodu at numerimoire.net Thu Jun 18 07:32:42 2009 From: nicolas.brodu at numerimoire.net (Nicolas Brodu) Date: Thu, 18 Jun 2009 09:32:42 +0200 Subject: [Causality-ML] Reminder: talk today (updated): On Granger Causality and Transcriptional Networks Message-ID: <200906180932.42563.nicolas.brodu@numerimoire.net> Dear Causality and Machine Learning group, This is a friendly reminder for today's talk by Radhakrishnan Nagarajan: "On Granger Causality and Transcriptional Networks" (title updated). When: Thursday 18 June, Paris 17h, ET 11h, PT 8h, Tokyo 0h(Friday) URL: http://www.afia-france.org/tiki-index.php?page=GroupeDeLecture090618RN Phone number: +1 (218) 936-7999 Participant code: 665140# Reference paper: ---------------- Nagarajan, R (2009): A Note on inferring acyclic network structures using Granger causality tests. The International Journal of Biostatistics. 5(1), Article 10. Abstract of the presentation: ----------------------------- Granger causality (GC) and its extension have been used widely to infer causal relationships from multivariate time series generated from biological systems. GC is ideally suited for causal inference in bivariate vector autoregressive process (VAR). A zero magnitude of the upper or lower off-diagonal element(s) in a bivariate VAR is indicative of lack of causal relationship in that direction resulting in true acyclic structures. However, in experimental settings, statistical tests, such as F-test that rely on the ratio of the mean-squared forecast errors, are used to infer significant GC relationships. The presentation will investigate acyclic approximations within the context of bi-directional two-gene network motifs modeled as bivariate VAR. The fine interplay between the model parameters in the bivariate VAR, namely: (i) transcriptional noise variance, (ii) autoregulatory feedback, and (iii) transcriptional coupling strength that can give rise to discrepancies in the ratio of the mean-squared forecast error are elucidated. Subsequently, their impact on statistical power is investigated using Monte Carlo simulations. While the results presented are generic, the presentation is likely to have significant impact on inferring functional relationships and transcriptional circuitry from gene expression profiles. ----------------------------- 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 planning for the next presentation is maintained at: http://www.afia-france.org/tiki-index.php?page=Groupe+de+lecture Tip for a cheap voice-over-IP connection to the conference: Open an account on nonoh.net (server in Germany) or voipcheap.com (server in US) or another SIP provider near your country. Pay 10 euros for 3 or 4 month unlimited access to nearly all land line phone numbers in the world, including the teleconference number. This means you will be able to spend as much time as you want for free on long-distance and international calls with the basic fee. Use a microphone + headset on your computer, and a program like Twinkle (Linux) or Ekiga (multi-OS) or X-Lite (windows). Set it up with your nonoh.net or voipcheap account (see the help page on these sites). Beware of filtering firewalls. Then call our teleconference number from your computer. Note: We are not affiliated to any of these groups. We provide this information in the hope it is useful. Best regards, Nicolas Brodu From nicolas.brodu at numerimoire.net Sat Jun 20 21:05:20 2009 From: nicolas.brodu at numerimoire.net (Nicolas Brodu) Date: Sat, 20 Jun 2009 23:05:20 +0200 Subject: [Causality-ML] Video replay, next presentations in September Message-ID: <200906202305.20936.nicolas.brodu@numerimoire.net> Dear Causality and Machine Learning group, The video of Radhakrishnan Nagarajan's talk on Granger Causality and Transcriptional Networks is now available online and for download: http://www.encours.org/causality/acycnet_gc/replay.html We are taking a summer break, and the presentations will start again in september. The mailing-list is of course still open for discussions, and we invite you to comment on past presentations. If you know of potentially interested speakers or if you wish to present a paper, please send us a message so we can plan ahead the next sessions. Thanks for your interest, Nicolas Brodu