pyMVPA Workshop 2015: Multivariate Pattern Analysis (MVPA) of Neuroimaging Data with PyMVPA

 Workshop

Multivariate pattern analysis (MVPA) approaches have gained great importance in cognitive neuroscience in recent years. In this two-day workshop we will explore the basic concepts of MVPA and its practical application to neuroimaging data. The course will be taught by the authors of PyMVPA, Michael Hanke (University of Magdeburg, Germany) and Yaroslav Halchenko (Dartmouth College, NH, USA). 

The workshop will feature lectures covering topics from basic principles of MVPA to various specific methodologies currently found in the literature. The majority of the time will be spent on hands-on sessions, in which participants will be introduced to PyMVPA (http://www.pymvpa.org), a free software package providing all tools necessary to conduct MVPA. During the hands-on sessions, step-by-step tutorials will guide through all steps of MVPA. An updated version of the extensive PyMVPA web-tutorial (http://www.pymvpa.org/tutorial.html) will serve as the base course material for the practical sessions. New topics will include a command line interface to PyMVPA with minimal requirements of Python programming skills and well as an introduction to the increasingly popular representational similarity analysis. 

Prerequisites - a basic understanding of Python is necessary for this workshop. For those wishing to attend, with no previous Python experience, online tutorials such as http://www.codecademy.com/en/tracks/python are advised.

Participants will be provided with computers, software and example datasets on site. In addition, course materials and data will be freely available for download after the end of the workshop.

The workshop will be focused on fMRI data analysis, but the majority of the covered strategies are also applicable to other data modalities such as EEG and MEG, and examples will be given.

Registration

Registration to this event is open from 16th February, and will remain open until all spaces are filled.  Please be aware that the number of places is limited by the size of our computer room and hands-on workshop facilities. Applications will be accepted on a first-come, first-served basis. 

The registration fee is £100. This will help cover printing, travel expenses for the speakers, and include coffee, tea and lunch on both days. 

To register for the workshop, please email cem552@york.ac.uk with your name and institution.

Accommodation

Participants will be responsible for making their own accommodation arrangements. The following website may be helpful:

http://www.yorkconferences.com/events/venue/bed-and-breakfast-14.aspx 

Preliminary Schedule

Day 1

• Administrative remarks

• Lecture ”A very short introduction to multivariate pattern analysis (MVPA) for neuroscience”

• Quiz: ”Python foundations”

• Hands-on “Data representation in PyMVPA”

• Hands-on “PyMVPA building blocks and the command line interface”

• Lecture ”Basic MVPA strategies”

• Hands-on “Classification and cross-validation“
  
  
• Hands-on “Meta-classifiers and Searchlights“

Day 2

• Lecture “Advanced methods, other data modalities and current developments”

• Hands-on “Searchlights (revamped) and RSA”

• Hands-on “Feature extraction and preprocessing”

• Lecture “PyMVPA and the larger scientific software eco-system”

• Hands-on “Group analyses”

• Hands-on “Statistical evaluation and Q&A”

   

   

  Contact Information 

  If you have any any enquiries about the pyMVPA workshop at the University of York, please email us at cem552@york.ac.uk

   

  Organisers: Dr Shirley-Ann Rueschemeyer and Charlotte Murphy