Sta­tis­ti­cal mo­del­ling

Ge­ne­ra­li­sed li­ne­ar mo­dels (GLMs)

GLMs ge­ne­ra­li­se li­ne­ar mo­dels by al­lo­wing for va­rious types of data (e.g., non-​normal data) and fle­xi­ble forms of the li­ne­ar pre­dic­tor. In this cour­se, we will dis­cuss va­rious aspects of GLMs, in­clu­ding the for­mu­la­ti­on of GLMs, model fit­ting, model selec­tion, and model che­cking. In ad­di­ti­on, we will cover va­rious ex­ten­si­ons of GLMs, in­clu­ding ge­ne­ra­li­sed li­ne­ar mixed mo­dels (GLMMs), ge­ne­ra­li­sed ad­di­ti­ve mo­dels (GAMs), and GAMs for lo­ca­ti­on, scale, and shape (GAMLSS).

Stu­dent feed­back can be found here.

Su­per­vi­si­on of B.Sc. and M.Sc. the­ses

If you are in­te­rested in wri­ting your B.Sc. or M.Sc. the­sis wit­hin sta­tis­ti­cal mo­del­ling, plea­se con­tact us. Pre­re­qui­si­te is the suc­cess­ful com­ple­ti­on of the mo­du­les 31-M3 Sta­tis­tics and 31-M9 Data ana­ly­sis. You can eit­her sug­gest a topic of your choice or ask us for ideas. Ap­p­lied the­ses in col­la­bo­ra­ti­on with com­pa­nies are also pos­si­ble. How­e­ver, note that (es­pe­cial­ly M.Sc.) the­ses should have a strong me­tho­do­lo­gi­cal com­po­nent. Use­ful tem­pla­tes (for both Word and LaTeX) are pro­vi­ded by Prof. Dr. Diet­mar Bauer.