Overview of TALENT mini-projects#
The exercises in this section are multi-part but are not full-blown projects, hence the name “mini-projects”. They originated with the TALENT (“Training in Advanced Low-Energy Nuclear Theory”) course given by the authors in 2019.
The contents are:
Mini-project I: Parameter estimation for a toy model of an EFT. The goal is to reproduce Bayesian parameter estimation results for a (published) toy model ofeffective field theories, which takes the form of Taylor series of a specified function.
Mini-project IIa: Model selection basics and Mini-project IIb: How many lines are there? explore model selection via evidence calculation, using as prototypical problems the number of terms in a Taylor series that provide the best to some (noisy) data in a finite range and inferring how many signal peaks there are in a noisy set of data.
Mini-project IIIa: Bayesian Optimization explores using Bayesian optimization to minimize functions of one variable and more variables while Mini-project IIIb: Bayesian Neural Networks explores a basic classification taks with a Bayesian neural network.