JOB #1 For September 13th: Chapter 3 of W&H. Read and digest. Exercises 3.6, 3.10 and 3.11. Here 3.10 is the big one. For this you will Develop code -- (Matlab preferably, R/Splus is OK, C/C++ OK too) that is modular, short and simple -- implementing the sequential learning in a general dynamic regression DLM, rather than just the one variable/no intercept model. The general model and its updating equations are the special case of the general DLM with Gt set to the identity matrix: read this off the tabular summary in Section 4.6 with that simplification. Implement it with the discount factor specification for evolution variances in which Wt = Ct-1(δ-1-1) For Exercise 3.10, you will simply apply this code in the special, simple case of one predictor/regressor and zero intercept. Use this to address exercise 3.10: This essentially just reproduces/refines the worked example of section 3.4.2. Include in your write-up your versions of figures 3.5-3.9. Hand in detailed code print-out, the figures above and comments on your experiences. Use Latex/Tex please. Sept 7th class: Carlos Carvalho will hold office hours throughout the class period in 025 Old Chem, for discussion of Chapter 3 and the big-picture issues as well as details. Hand-in completed assignment (in Mike W's Duke Statistics mailbox in room 211 Old Chem) before 5pm on Tuesday 13th September.