This may be of interest to some of you. The punchline for Bayesians is "don't make point estimates". <br><br><div class="gmail_quote">---------- Forwarded message ----------<br>From: <b class="gmail_sendername">Marta Kutas</b> <span dir="ltr"><<a href="mailto:mkutas@ucsd.edu">mkutas@ucsd.edu</a>></span><br>
Date: Fri, Nov 13, 2009 at 2:25 PM<br>Subject: next kutas lab meeting<br>To: <a href="mailto:department@cogsci.ucsd.edu">department@cogsci.ucsd.edu</a>, Esmeralda De Ochoa <<a href="mailto:edeochoa@cogsci.ucsd.edu">edeochoa@cogsci.ucsd.edu</a>>, Paul Krewski <<a href="mailto:pkrewski@cogsci.ucsd.edu">pkrewski@cogsci.ucsd.edu</a>>, David Groppe <<a href="mailto:dgroppe@cogsci.ucsd.edu">dgroppe@cogsci.ucsd.edu</a>>, Ben Cipollini <<a href="mailto:bcipolli@cogsci.ucsd.edu">bcipolli@cogsci.ucsd.edu</a>>, Vicente Malave <<a href="mailto:vicente.malave@gmail.com">vicente.malave@gmail.com</a>>, Paul Hammon <<a href="mailto:paul.hammon@gmail.com">paul.hammon@gmail.com</a>>, Nathan White <<a href="mailto:natew4@yahoo.com">natew4@yahoo.com</a>>, Marta Kutas <<a href="mailto:kutas@cogsci.ucsd.edu">kutas@cogsci.ucsd.edu</a>>, Steve Barrera <<a href="mailto:sjbarrera@ucsd.edu">sjbarrera@ucsd.edu</a>>, Irina Gorodnitsky <<a href="mailto:igorodni@cogsci.ucsd.edu">igorodni@cogsci.ucsd.edu</a>>, "Leslie J. Carver" <<a href="mailto:ljcarver@ucsd.edu">ljcarver@ucsd.edu</a>>, "Steven A. Hillyard" <<a href="mailto:sah@sdepl.ucsd.edu">sah@sdepl.ucsd.edu</a>>, Seana Coulson <<a href="mailto:coulson@cogsci.ucsd.edu">coulson@cogsci.ucsd.edu</a>>, Aniruddh Patel <<a href="mailto:apatel@nsi.edu">apatel@nsi.edu</a>>, Eric Halgren <<a href="mailto:ehalgren@ucsd.edu">ehalgren@ucsd.edu</a>>, "Anders M. Dale" <<a href="mailto:amdale@ucsd.edu">amdale@ucsd.edu</a>>, Robert Kluender <<a href="mailto:kluender@ling.ucsd.edu">kluender@ling.ucsd.edu</a>>, Ying Wu <<a href="mailto:ywu@cogsci.ucsd.edu">ywu@cogsci.ucsd.edu</a>>, David Brang <<a href="mailto:dbrang@ucsd.edu">dbrang@ucsd.edu</a>>, Jennifer Collins <<a href="mailto:jcollins@cogsci.ucsd.edu">jcollins@cogsci.ucsd.edu</a>>, Tristan Davenport <<a href="mailto:punninglinguist@gmail.com">punninglinguist@gmail.com</a>>, Julia Evans <<a href="mailto:jevans@mail.sdsu.edu">jevans@mail.sdsu.edu</a>>, Marguerite Mcquire <<a href="mailto:mmcquire@crl.ucsd.edu">mmcquire@crl.ucsd.edu</a>>, Matt Schalles <<a href="mailto:matthewdaves@yahoo.com">matthewdaves@yahoo.com</a>>, Melanie Tumlin <<a href="mailto:melanie.tumlin@gmail.com">melanie.tumlin@gmail.com</a>>, Laura Kertz <<a href="mailto:kertz@ling.ucsd.edu">kertz@ling.ucsd.edu</a>>, "Anna E. M. Holt" <<a href="mailto:aeholt@cogsci.ucsd.edu">aeholt@cogsci.ucsd.edu</a>>, Chris Barkley <<a href="mailto:cbarkley@ling.ucsd.edu">cbarkley@ling.ucsd.edu</a>>, lrosenfelt <<a href="mailto:lrosenfelt@ling.ucsd.edu">lrosenfelt@ling.ucsd.edu</a>>, Klinton Bicknell <<a href="mailto:kbicknell@ucsd.edu">kbicknell@ucsd.edu</a>>, Wen-hsuan Chan <<a href="mailto:whchan@cogsci.ucsd.edu">whchan@cogsci.ucsd.edu</a>>, Philip Hofmeister <<a href="mailto:phofmeister@ucsd.edu">phofmeister@ucsd.edu</a>>, Lindsay Crissman <<a href="mailto:lcrissman@cogsci.ucsd.edu">lcrissman@cogsci.ucsd.edu</a>>, Nathaniel Smith <<a href="mailto:njsmith@cogsci.ucsd.edu">njsmith@cogsci.ucsd.edu</a>>, Marisa Sizemore <<a href="mailto:msizemore@crl.ucsd.edu">msizemore@crl.ucsd.edu</a>>, Ross Metusalem <<a href="mailto:rmetusal@cogsci.ucsd.edu">rmetusal@cogsci.ucsd.edu</a>>, Jamie Alexandre <<a href="mailto:jdalexan@ucsd.edu">jdalexan@ucsd.edu</a>>, Kim Sweeney <<a href="mailto:ksweeney@cogsci.ucsd.edu">ksweeney@cogsci.ucsd.edu</a>><br>
<br><br><br>
Monday, Nov 16, 3:30, CSB 280<br>
<br>
Vicente Malave<br>
<br>
Classification Caveats : Practical limits of Multi-Voxel Pattern Analysis in fMRI.<br>
<br>
Multivariate pattern classification techniques are attracting increasing attention as an alternative data analysis technique in cognitive neuroscience. The purpose of this talk is to provide a critical look at two inferences using these methods. The first inference is simply that two conditions of the experiment are different. We discuss the connection between pattern classification methods and classical multivariate hypothesis testing. We briefly discuss resampling-based statistics, statistical power, and the impact on experimental design. The second inference is in interpretation of the model learned by a linear classifier. In particular, many studies compute "importance maps" that purport to show where discriminating information between the two conditions is located in the brain. We provide a critical look at this inference by noting that many very different models can achieve similar classification accuracy, and many very similar models can perform at chance. We show some techniques for exploring the space of possible models, discuss proposed solutions from the literature, and urge caution when making inferences about the importance of particular feature in classification.<br>
<br>
<br>
all welcome<br>
</div><br><br clear="all"><br>-- <br>Vicente Malave<br>