ABS08 - 2008 Applied Bayesian Statistics School

BAYESIAN DECISION PROBLEMS IN BIOSTATISTICS AND CLINICAL TRIALS

Centro Congressi Panorama, Trento, Italy

Monday June 9

· [08.30-         ] Registration

· [09.00-10.30] I - Review of Bayesian statistical inference

1.      Bayesian inference

2.      Priors

3.      Posterior asymptotics

· [10.30-11.00] Break

· [11.00-12.30] I - Review of Bayesian statistical inference

4.      Model comparison

5.      Bayesian decision problems

· [14.00-15.30] II - Principles of clinical trial design

6.      Clinical trial design

7.      Hierarchical models

           · [15.30-16.00] Break

· [16.00-17.30] Practical Session 1: Basics of R, packages, MCMCpack

· [19.00-21.00] Welcome buffet (at Centro Congressi Panorama)

Tuesday June 10

                            Sequential Decisions

· [09.00-10.30] III - Dose finding studies -- simulation based sequential design

8.      Bayesian dose finding

9.      Bayesian decision theoretic adaptive dose allocation

· [10.30-11.00] Break

· [11.00-12.30] III - Dose finding studies -- simulation based sequential design

10. Sequential stopping

· [14.00-15.30] IV - Drug screening designs -- decision boundaries

11. Screening design

12. Decision boundaries

13. Uncertainty and validation

· [15.30-16.00] Break

· [16.00-17.30] Practical Session 2: clinical trial simulation, operating characteristics

Wednesday June 11

· [09.00-10.30] V - Myopic sequential designs and randomization

14. Setup and notation

15. Expected and optimal utility

16. Randomization: nondominated actions

17. Example: PEG Intron trial

· [10.30-11.00] Break

                    Multiple comparisons

· [11.00-13.00] VI - Microarray group comparison

18. Hierarchical mixture models

19. Gamma/Gamma model

20. POE

· [13.00-         ] Free time

Thursday June 12

· [09.00-10.30] VII - Multiplicities, error rates and controls

21. FDR, Bayesian FDR and local FDR

22. Posterior adjustment for multiplicities

23. FDR and Bayesian decision theory

24. Comments on BH

· [10.30-11.00] Break

· [11.00-12.30] VIII - Optimal discovery procedure

25. Optimal discovery procedure

26. Bayesian discovery procedure

· [14.00-15.30] Participants' presentations

·        Francesca Ieva, Politecnico di Milano, Italy

      The Milano network for acute coronary syndromes and emergency services: a biostatistical case study

·        Valeria Vitelli, Politecnico di Milano, Italy

      k-means clustering and principal points: an open problem on functional data

·        Tobias Sing

Visualizing classifier performance in R using ROCR

· [15.30-16.00] Break

· [16.00-17.30] Practical Session 3: POE and EBarrays packages

· [20.30-         ] Farewell dinner (Osteria San Rocco in Sardagna)

Friday June 13

                    ROC Curves

· [09.00-10.30] IX - Bayesian inference for ROC curves

27. Bayesian inference for ROC curves

28. Semiparametric Bayesian ROC analysis

· [10.30-11.00] Break

· [11.00-13.00] X - Random discontinuation designs