The field of statistics has undergone profound changes in recent decades. Firstly, the types of datasets that statisticians are asked to analyse have transformed dramatically. In the past, we typically dealt with datasets containing many observations and a modest number of carefully chosen variables. Today, by contrast, it is common to encounter datasets with thousands of variables---sometimes even far exceeding the number of observations. For instance, in genomics, we might measure the expression levels of several thousand genes but only across a few hundred tissue samples. Classical statistical methods are often simply not applicable in these ``high-dimensional'' settings. As the scale of data collection has expanded, so too has the scope of the questions we seek to answer. Whereas statistics was once primarily concerned with uncovering associations between variables, we are now increasingly interested in understanding the causal structure of data. And rather than focusing solely on prediction, we often aim to predict the effects of interventions. At the same time, the rapid rise of machine learning has provided us with powerful new tools. In this course, we will explore how these advances can be harnessed to tackle some of the modern statistical challenges outlined above. The selection of material is heavily biased towards my own interests, but I hope it will nevertheless give you a flavour of some of the most important recent methodological developments in statistics.
The code for the demonstrations is written in R. Rstudio is a useful editor for R. Here are some introductory worksheets on R: Sheet 1, (solutions); Sheet 2, (solutions). The code for the demonstrations is given below.