xuerui wang, rebecca hutchinson and tom m. mitchell- training fmri classifiers to detect cognitive...

8

Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects Xuerui Wang, Rebecca Hutchinson, and Tom M. Mitchell Center for Automated Learning and Discovery Carnegie Mellon Universit y 5000 Forbes Av enue, Pittsburgh, PA 15213 xuerui.wang, rebecca.hutchinson, tom.mitchell @cs.cmu.edu Abstract We consi der lear ning to clas sify cogni tiv e states of human subject s, bas ed on the ir bra in act iv ity obs erved via functi onal Magnetic Res ona nce Imaging (fMRI). This problem is important because such classifiers con- stitute “virtual sensors” of hidden cognitive states, which may be useful in cognitive sci ence research and clinical applicati ons. In recent work, Mitchell, et al. [6,7,9] have demonstrated the feasibility of training such classifiers for individual human subjects (e.g., to distinguish whether the subject is reading an ambiguous or unambiguous sentence, or whether they are reading a noun or a verb). Here we extend that line of research, exploring how to train classifiers that can be applied across multiple human subjects, including subjects who were not involved in training the clas sifie r. W e describe the desi gn of severalmachi ne learn ing appro aches to training multiple-subject classifiers, and report experimental results demonstrating the success of these methods in learning cross-subject classifiers for two different fMRI data sets. 1 Int roduc tio n The advent of functional Magnetic Resonance Imaging (fMRI) has made it possible to safely , non-invasiv ely observe correlates of neural activity across the entire human brain at high spatial resolution. A typical fMRI session can produce a three dimensional image of brain activation once per second, with a spatial resolution of a few millimeters, yielding tens of millions of individual fMRI observations over the course of a twenty minute session. This fMRI technology holds the potential to revolut ionize studies of human cogniti ve processing, provided we can develop appropriate data analysis methods. Researchers have now employed fMRI to conduct hundreds of studies that identify which regions of the brain are activated on average when a human performs a particular cognitive task (e.g., reading, puzzle solving). Typical research publications describe summary statis- tics of brain activity in various locations, calculated by averaging together fMRI observations collected over multiple time intervals during which the subject responds to repeated stimuli of a particular type. Our interest here is in a different problem: training classifiers to automatically decode the

Upload: noscript

Post on 06-Apr-2018

231 views

Category:

Documents

0 download

Report

Download

Embed Size (px):

TRANSCRIPT

Page 1: Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

8/3/2019 Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

http://slidepdf.com/reader/full/xuerui-wang-rebecca-hutchinson-and-tom-m-mitchell-training-fmri-classifiers 1/8

Page 2: Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

8/3/2019 Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

http://slidepdf.com/reader/full/xuerui-wang-rebecca-hutchinson-and-tom-m-mitchell-training-fmri-classifiers 2/8

Page 3: Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

8/3/2019 Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

http://slidepdf.com/reader/full/xuerui-wang-rebecca-hutchinson-and-tom-m-mitchell-training-fmri-classifiers 3/8

Page 4: Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

8/3/2019 Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

http://slidepdf.com/reader/full/xuerui-wang-rebecca-hutchinson-and-tom-m-mitchell-training-fmri-classifiers 4/8

Page 5: Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

8/3/2019 Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

http://slidepdf.com/reader/full/xuerui-wang-rebecca-hutchinson-and-tom-m-mitchell-training-fmri-classifiers 5/8

Page 6: Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

8/3/2019 Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

http://slidepdf.com/reader/full/xuerui-wang-rebecca-hutchinson-and-tom-m-mitchell-training-fmri-classifiers 6/8

Page 7: Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

8/3/2019 Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

http://slidepdf.com/reader/full/xuerui-wang-rebecca-hutchinson-and-tom-m-mitchell-training-fmri-classifiers 7/8

Page 8: Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

8/3/2019 Xuerui Wang, Rebecca Hutchinson and Tom M. Mitchell- Training fMRI Classifiers to Detect Cognitive States across Multiple Human Subjects

http://slidepdf.com/reader/full/xuerui-wang-rebecca-hutchinson-and-tom-m-mitchell-training-fmri-classifiers 8/8

Advisor: Jun Zhu Speaker: Xuerui Wang

Decision Tree Classifiers

Commonly Used Classifiers

Perceptron Linear Classifiers

Classifiers in Kam-Tai Languages · 2013-01-23 · 4.2.7. Classifiers and temporal words ..... 169 4.2.8. Classifiers and pronouns ..... 171 4.2.9. Classifiers and numerals ... few

MACHINE LEARNING Classifiers: Support Vector Machinelasa.epfl.ch/teaching/lectures/ML_Phd/Slides/SVM.pdfMACHINE LEARNING 4 Classifiers There is a plethora of classifiers, e.g: - Neural

Michael Hutchinson

1 Classifying Instantaneous Cognitive States from fMRI Data Tom Mitchell, Rebecca Hutchinson, Marcel Just, Stefan Niculescu, Francisco Pereira, Xuerui

8. Classifiers - MPG.PuRe

Hutchinson powerpoint

Ideology Classifiers for Political Speech Stefan Kaufmannccc.inaoep.mx/~villasen/bib/Ideology Classifiers for Political Speech.pdf · Ideology classifiers for political research pg

Dynamic Selection of Classifiers - UFPRSelection of classifiers A single or an ensemble of classifiers can be selected. Static: performed during training, the same selected classifiers

Grande Day Program - Hutchinson Jaycee Water … Hutchinson Marching Tigers Hutchinson Pet Hospital, Dr. Curtis Reiter 14 Independent School District #423 AKI 15 2009 Hutchinson Jaycee

Linear Classifiers/SVMs

Lecture8 classifiers ldc_rules

Learning Bayesian Classifiers

Bayes Classifiers

Evaluation Metrics (Classifiers)

Naïve Bayes Classifiers

Evaluating Classifiers

Hutchinson Public Schools - Hutchinson High School

5 character classifiers

Advanced classifiers

Personalized classifiers

Lecture 12 Classifiers Part 2 Topics Classifiers Maxent Classifiers Maximum Entropy Markov Models Information Extraction and chunking intro Readings: Chapter

classifiers kelompok 1

Hutchinson group

Nearest Neighbourhood Classifiers in Biometric Fusion · Nearest Neighbourhood Classifiers in Biometric Fusion Nearest Neighbourhood Classifiers in Biometric Fusion ... Electronic

Separating and Classifiers

Gaussian Classifiers

Gaussian Bayes Classifiers

8. Classifiers - Max Planck Societypubman.mpdl.mpg.de/.../component/escidoc:1582926/Classifiers.pdf · 8. Classifiers 1. Introduction 2 ... and examples of classifiers in various

Speaker: Xuerui Wang Advisor : Jun Zhu Dec. 27. 2013

Theoretical Study on the silaaromatics Speaker: Xuerui Wang Advisor : Jun Zhu Dec. 23. 2013