PhD Courses in Denmark

A student who has met the objectives of the course will be able to conduct a custom Bayesian statstical analysis of realistic biological data using the latest available theoretical and computational methods.

A student who has met the objectives of the course will be able to:

• Beskriv Bayesiansk inferens abstrakt
• Vurder om Bayesiansk inferens er en god løsning på et problem
• Formuler brugerdefinerede målemodeller til at beskrive biologiske problemer
• Løs statistiske modelleringsproblemer ved iterativ tilpasning og evaluering af en række modeller
• Vælg passende software til et Bayesiansk statistisk modelleringsprojekt
• Forstå gradientbaserede MCMC-teknikker og deres fejltilstande
• Tilpas biologiske modeller med indlejrede ODE-systemer, rodfindingsproblemer og Gaussiske processer.
• Udfør Bayesiansk optimering
• Forstå de seneste tendenser inden for Bayesiansk statistisk inferens

Day 1. am: Bayesian statistical inference, motivating example pm: Set up computers (Python, uv, git, editor) Day 2. am: Regression, formula-based models and why they aren't enough pm: Some regression examples, bambi Day 3. am: Markov Chain Monte Carlo, why you still probably want to use it. pm: A Bayesian statistics stack for computational biology Day 4. am: What to do with MCMC output? pm: Worked examples: - convergence - divergent transitions - model comparison - change of variables causing bad model Day 5. am: Bayesian workflow pm: Workflow example with automation Day 6. am: Ordinary differential equations pm: Diffrax, fermentation examples Day 7. am: Algebraic equation systems, implicit differentiation pm: Optimistix, steady state example, grapevine Day 8. am: Gaussian processes, HSGPs pm: GP example Day 9. am: Bayesian optimisation pm: BO example Day 10. am: Fun new Bayesian trends - Probabilistic numerics - Amortised Bayesian inference - New MCMC algorithms - Control - Normalising flows pm: examples Day 11-15: Supervised project