BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:IC Colloquium : Bayesian methodologies for efficient data analysis
DTSTART:20160201T101500
DTEND:20160201T113000
DTSTAMP:20260407T110531Z
UID:c7c9562aafff13642905ed028d296dad0435fe947966a457c576fcc3
CATEGORIES:Conferences - Seminars
DESCRIPTION:By : Mijung Park - University of Amsterdam\nIC Faculty candida
 teAbstract :\nMachine learning and data science can greatly benefit from B
 ayesian methodologies\, not only because they improve generalisation perfo
 rmance compared to point estimates that are prone to overfitting\, but als
 o they provide efficient and principled ways to solve a broad range of sta
 tistical problems. In this talk\, I will describe several concrete example
 s where using Bayesian approaches greatly benefit in tackling problems occ
 urring in many areas of science. These examples include (a) designing prio
 rs using domain knowledge for structurally sparse high-dimensional paramet
 ers with application to functional neuroimaging data and neural spike data
 \; (b) Bayesian manifold learning that enables evaluating the quality of e
 stimated latent manifold as well as learning the latent dimension from sta
 tistical evidence\; and (c) approximate Bayesian computation (ABC) for mod
 els with intractable likelihoods\, where we employ kernel mean embeddings 
 to measure data similarities\, which is an essential step in ABC.Bio :\nMi
 jung Park completed her PhD in the department of Electrical and Computer E
 ngineering under the supervision of Prof. Jonathan Pillow and Prof. Al Bov
 ik at The University of Texas at Austin. She was a postdoctoral research f
 ellow working with Prof. Maneesh Sahani at the Gatsby computational neuros
 cience unit at University College London. Currently\, she is a postdoctora
 l research fellow working with Prof. Max Welling in the informatics instit
 ute at University of Amsterdam.More information
LOCATION:BC 420 https://plan.epfl.ch/?room==BC%20420
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR