Full list of models tested.

Summary of performance for 1061 formulations of the linear STRF (see Fig 2), averaged across the N = 176 A1 vocalization datasets. Each row indicates parameter count, fit performance (estimation data), test performance (validation data), and test performance after correcting for validation sampling limitations (Eq 23) for a single model. The first 163 rows describe models detailed in this study, and the remainder are additional suboptimal models that were tested. The key on the first two pages indicates how to interpret model names.

(PDF)