## General Linear Modelling

### Linear Algebra

Like many scientific fields, neuroimaging relies enormously on linear algebra and it is thus worthwhile investing in learning it. Beside this application, all the usual statistics relies on the General LInear Model or related, and this means if you understand linear algebra, you already understand the modelling and pretty much the inferential part of statistical testing. To get you started I have some old material: an introduction to vectors to understand the relationship between geometry and algebra and a short text reviewing basic matrix operations.

Check also the MIT excellent linear algebra course, for free!

### General Linear Modelling

This Matlab GLM page explains what is a linear system and how to solve simple and multiple regressions and ANOVA. All of this is performed using standard linear algebra, i.e. matrix inversion. Everything you always wanted to ask about linear independence, orthogonalitity and correlation .. in a bit more than a one page article. I also wrote something on what makes a contrast orthogonal.

## Robust Statistics

Ordinary Least Squares are not robust because the mean has a 0 breakdown point. Many alternartives exist and I use and develop such methods, in particular robust correlations and tests based on robust estimators.

## Digital Signal Processing

### Fourrier analyses

Fourrier analyses are usuful in so many applications, it is worth knowking about it: here is a pdf with code that explains how Fourrier analyses work and things you can do