Das Berliner Kolloquium „Statistische Methoden in der empirischen Forschung“ lädt ein zur Veranstaltungsreihe des Wintersemesters 21/22, die wöchentlich dienstags um 17:00 stattfindet. Start war 19. Oktober 2021.
Am 23.11.2021 um 17:00 bis ca. 18:30: "Random Forests: Myths and Facts" mit Marvin N. Wright.
Abstract: Random forests rarely win prediction challenges. However, if implementation time and/or computing time are taken into account they are hard to beat. Random forests work well without tuning and on high dimensional data and they are interpretable. These and other claims are very common in the literature about random forests, but are they facts or just myths? In this talk, I will analyze common claims about random forests and evaluate their degree of truth.
Einwahllink: https://beuth-hochschule-de.zoom.us/j/83346251825?pwd=MTZQc3RTRGRQL2dkeExXR251b1FiUT09
