INB Lunch-Seminar

Inductive ambiguity and the theorem of the ugly duckling

PD Dr.-Ing. Erhardt Barth

Machine learning techniques have a wide and growing spectrum of applications aided by advancements in both the theoretical foundations and the available computational power. However, in the 50s, 60s, and 70s some pioneers of the field computed little but wrote comprehensive textbooks (with typewriters). In order to minimize the risk of reinventing the wheel, and especially the flat tire, we want to discuss some of this early work. In this seminar we will briefly review the book “Pattern Recognition: Human and Mechanical” by Satosi Watanabe, an IBM scientist and professor at the universities of Yale and Hawaii. The focus will be on his Theorem of the Ugly Duckling, which demonstrates that classification is impossible without some sort of bias.

Zeit: Freitag, den 20. April 2007, 12 Uhr c.t.

Ort: Institut für Neuro- und Bioinformatik
        Seminarraum (1. OG, Raum 17),
        Ratzeburger Allee 160 (Geb. 64, 1. OG)