BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:IC Colloquium: Deep Learning for Network Biomedicine
DTSTART:20190304T101500
DTEND:20190304T111500
DTSTAMP:20260406T063917Z
UID:30fd528c687f48c04a17329ee0ec47eb056a8e692cdc7ed7ae0925dc
CATEGORIES:Conferences - Seminars
DESCRIPTION:By: Marinka Zitnik - Stanford University\nIC Faculty candidate
 \n\nAbstract:\nLarge datasets are being generated that can transform biolo
 gy and medicine. New machine learning methods are necessary to unlock thes
 e data and open doors for scientific discoveries. In this talk\, I will ar
 gue that\, in order to advance science\, machine learning models should no
 t be trained in the context of one particular dataset. Instead\, we should
  be developing methods that can integrate rich\, heterogeneous data and kn
 owledge into multimodal networks\, enhance these networks to reduce biases
  and uncertainty\, and learn over the networks. My talk will focus on two 
 key aspects of this goal: deep learning and network science for multimodal
  networks. I will first show how we can move beyond prevailing deep learni
 ng methods\, which treat network features as simple variables and ignore i
 nteractions between entities. Further\, I will present an algorithm that l
 earns deep models by embedding multimodal networks into compact embedding 
 spaces whose geometry is optimized to reflect the interactions\, the essen
 ce of multimodal networks. These deep models set sights on new frontiers\,
  including the prediction of protein functions in specific human tissues\,
  modeling of drug combinations\, and repurposing of old drugs for new dise
 ases. Beyond such predictive ability\, a hallmark of science is to achieve
  a holistic understanding of the world. I will discuss how we can blend ne
 twork algorithms with rigorous statistics to harness biomedical networks a
 t the scale of billions of interactions. These methods revealed\, among ot
 hers\, how Darwinian evolution changes molecular networks\, providing evid
 ence for a longstanding hypothesis in biology. In all studies\, I collabor
 ated closely with experimental biologists and clinical scientists to give 
 insights and validate predictions made by our methods. I will conclude wit
 h future directions for contextual models of rich interaction data which o
 pen up new avenues for science.\n\nBio:\nMarinka Zitnik is a postdoc in Co
 mputer Science at Stanford University. Her research investigates machine l
 earning for biomedical sciences\, focusing on new methods for large networ
 ks of interactions between biomedical entities. Her methods have had a tan
 gible real-world impact in biology\, genomics\, and medicine\, and are use
 d by major biomedical institutions\, including Baylor College of Medicine\
 , Karolinska Institute\, Stanford Medical School\, and Massachusetts Gener
 al Hospital. She has multiple first-author papers in the top scientific jo
 urnals (PNAS\, Nature Communications) as well as in the top machine learni
 ng and computational biology venues (JMLR\, NIPS\, IEEE TPAMI\, KDD\, Bioi
 nformatics\, ISMB\, RECOMB). She received her Ph.D. in Computer Science fr
 om University of Ljubljana while also researching at Imperial College Lond
 on\, University of Toronto\, Baylor College of Medicine\, and Stanford Uni
 versity. Her work received several best paper\, poster\, and research awar
 ds from the International Society for Computational Biology. She was selec
 ted a Google Anita Borg Scholar\, Young Fellow at Heidelberg Laureate Foru
 m\, and received Jozef Stefan Golden Emblem Prize. In 2018\, she was named
  a Rising Star in EECS by MIT and also a Next Generation in Biomedicine by
  The Broad Institute of Harvard and MIT\, being the only young scientist w
 ho received such recognition in both EECS and Biomedicine. She is also a m
 ember of the Chan Zuckerberg Biohub at Stanford.\n\nMore information
LOCATION:BC 420 https://plan.epfl.ch/?room==BC%20420
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR