Contact Session #5
Statistical connectomics. Null models, permutations, machine learning.
Intended learning outcomes:
- Familiarize with concepts from non-parametric statistics applied to network measures and brain connectivity
- Consider the various methods used to compute statistical significance and control for multiple comparisons
- Understand how machine learning can be applied with brain connectivity
- Critically evaluate statistical results from the brain connectivity literature
Please read this paper very carefully. It is not too long but it is a very important one. Bring your questions/views/opinions to the lecture because we will spend some 20 minutes discussing it in the class all together.
Greenland, S., Senn, S. J., Rothman, K. J., Carlin, J. B., Poole, C., Goodman, S. N., & Altman, D. G. (2016). Statistical tests, P values, confidence intervals, and power: a guide to misinterpretations. European journal of epidemiology, 31(4), 337-350.
If some of you want to recommend other relevant readings on similar topics, please feel free to post them on the forum. Here some interesting ones
http://journals.sagepub.com/doi/pdf/10.1177/0146167217729162 and in general his blog: http://andrewgelman.com/
If you cannot access any of the papers above, ask in the forum and somebody will give you the PDF.
Lecture slides and material
Slides and materials will be uploaded at the bottom of this page.
If you couldn't make it to the first contact session, please catch up with the others with the assignment here below.