Topological complexity of smooth random functions. École d'Été de Probabilités de Saint-Flour XXXIX-2009. (Q630636)

These little lecture notes are a rare delight. The authors succeed in an impressive manner to combine a writing style that focuses on the main ideas and intuitions while still stating all the results in full mathematical rigor. They take the reader on an exciting journey through the theories of Gaussian processes and differential topology and geometry and then show how fascinating mathematics arises when combining these fields, not to speak about the wide range of applications. After an introduction that gives a glimpse on the forthcoming chapters, the authors discuss some basic material on Gaussian processes such as their construction, exceedance probabilities, regularity properties, local isotropy and the induced metric. They then move on to explain the necessary notions and results from differential geometry. They give most results for stratified manifolds. The main topics here are Lipschitz-Killing curvatures, Gaussian Minkowski functionals and kinematic formulae. The main result of the theoretical part of the text is the Gaussian kinematic formula for vector-valued random processes with iid centered and unit variance Gaussian components on regular stratified manifolds. The application part on topological inference uses examples from brain imaging and cosmology to illustrate what has been done in practice in recent years. Here, the Gaussian kinematic formula plays a major role to justify the so-called Euler characteristic heuristic. In the part on brain imaging, the authors explain how the Euler characteristic heuristic can be used to detect regions of increased brain activity in positron emission tomography, whereas the part on cosmology deals with the analysis of the cosmic microwave background radiation. The last chapter gives an outlook on future research on algebraic topology of excursion sets.