Larry Manevitz: “Computational Biomarkers , Machine Learning and Neurocomputation:Some recent work from the Neurocomputation Laboratory”

2015年07月02日  12:15 〜 13:00

CiNet 1F Conference Room

Larry Manevitz

Professor, Department of Computer Science, University of Haifa
Head, Neurocomputation Laboratory, Rothschild Institute, University of Haifa

Prof. Manevitz is in the Computer Science Department at the University of Haifa, Israel and is the head of the Neurocomputation Laboratory situated at the Rothschild Institute also at the University of Haifa. He is a mathematician and computer scientist, trained at Yale University and has worked in neurocomputation, brain and psychological neuroscience modeling, and artificial intelligence (especially in combining uncertain information). He has had visiting positions at Oxford University (Exp. Psychology), NYU (Courant Institute), NASA (Ames), U. of Texas, Hebrew University, amongst others. His laboratory web site is neurocomputation.wordpress.com

He will be visiting CiNet, Osaka thru August and is very interesting in hearing about the great work being performed here with an eye towards future collaborator work.

Abstract:

A biomarker can be defined as clear distinct physiological indicator of an abstract state, which could be a disease or a state of mind.   Traditionally biomarkers are relatively simple indicators (such as blood sugar level for diabetes) , but recent computational work shows that complex informational biomarkers can be discovered by machine learning techniques.

In this talk, we will try to illustrate how these strong computational methods can be used.  As time allows, some of the following examples taken from work performed in the Neurocomputation Laboratory in Haifa will be presented: 1) Early diagnosis of Parkinson Disease from Speech signals 2) Personalized BOLD signal modeling and brain mapping 3) Detection of a secondary declarative memory system in adult humans and (if time allows) 4) One class classification of cognitive visual tasks via neural networks, genetic algorithms and deep learning.