Statistical models in biology

All models are wrong, but some models are useful.
— George Box, British statistician

These courses address the principles of data analysis in biology. You may "bring your own data" to some of them.

Foundational courses

Start with these lectures to gain a solid understanding of the basics of statistical analysis in a biological context.

• Quick introduction to data analysis
• Basic statistics with Python
• Think Statistics with R
• ANOVA with R
• Bayesian statistics with R

Advanced courses

"Machine Learning" !
"Neural Networks" !!
"AI" !!!

Hype and buzzwords come and go. At the end, statistical models remain. We offer in-depth courses that teach you statistical model building in a biological context.

Regression techniques

The basic assumption behind these methodologies is that you observe one or more "responses" or "dependent variables" that are influenced by one ore more "predictors" or "independent variables". The observations themselves are noisy, but we assume that there is an underlying mathematical relationship between these variables. This is called the model.

The models usually have parameters. For instance a straight line can be specified by two parameters, the intercept and the slope. Fitting the model involves adjusting its parameters so that the model is "optimally close" (in some sense) to the training data. This procedure is called "learning" in ML and "training" in NN and "AI" circles.

• Linear regression with R
• Nonlinear regression with R
• Generalized linear models with R

Planned courses

• Clustering methods
• Classification methods
• Regression with artificial neural networks

Obsolete courses

These courses are not offered any longer.

• RNA-Seq data analysis
• Biocybernetics