External Lecturer:
Roberto Trotta
Course Type:
PhD Course
Academic Year:
2020-2021
Period:
January - February
Duration:
36 h
CFU (LM):
6
Description:
Each bloc requires 6 hrs; 24 lectures + 4 x 3 hrs labs = 36 hours (5 weeks: 11/01-24/02)
- Foundations of Bayesianism: Jaynes’ robot; Cox theorem, objective and subjective Bayes; prior choice (maximum entropy, conjugancy, Jeffreys’ prior, empirical Bayes, hyperpriors, etc); sensitivity analysis.
- Lab 1: Sensitivity analysis and volume effects.
- Advanced sampling methods: slice sampling, Langevin and Hamiltonian Monte Carlo, collapsed and augmented Gibbs sampling, reversible jump MC.
- Lab 2: Writing an MCMC from scratch.
- Bayesian model comparison: stopping rule paradox, p-values, Lindley paradox; Bayesian evidence, Bayes factor and interpretation. Savage-Dickey Density Ratio, Laplace approximation, Approximate Bayesian Computation (ABC).
- Lab 3: Applications of Bayesian model comparison.
- Bayesian model averaging, Bayesian optimization and experiment design.
- Lab 4: Bayesian optimization.
Please, notice that this is a course belonging to Data Science Excellence Department programme. MAMA PhD students can plan 33% of their credits (i.e. 50 hrs) from this programme.
Research Group:
Location:
A-128