A Replica Based Approach to Machine Learning
- Datum
- 08.08.2017
- Zeit
- 10:30 - 11:30
- Sprecher
- Dr. Zohar Nussinov
- Zugehörigkeit
- Washington University, St. Louis
- Sprache
- en
- Hauptthema
- Materialien
- Andere Themen
- Materialien, Physik
- Host
- Grit Rötzer
- Beschreibung
- We introduce a machine learning algorithm for classification and regression problems. The method is based on the invariances or stability of predicted results when known data is represented as expansions in terms of various stochastic functions. The few parameters (typically only one parameter is used in all studied examples) that this model has may be automatically optimized. The algorithm has been tested on data sets of various types and has been shown to consistently exhibit high accuracy and readily allow for optimization of parameters, while simultaneously avoiding pitfalls of existing algorithms such as those associated with class imbalance. Extensions of this method have been further applied to combinatorial optimization problems, image segmentation, and material prediction .
- Links
Letztmalig verändert: 08.08.2017, 09:53:29
Veranstaltungsort
Leibniz Institut für Festkörper- und Werkstoffforschung Dresden (B3E.26, IFW Dresden)Helmholtzstraße2001069Dresden
- Homepage
- http://www.ifw-dresden.de
Veranstalter
Leibniz Institut für Festkörper- und Werkstoffforschung DresdenHelmholtzstraße2001069Dresden
- Homepage
- http://www.ifw-dresden.de
Legende
- Ausgründung/Transfer
- Bauing., Architektur
- Biologie
- Chemie
- Elektro- u. Informationstechnik
- für Schüler:innen
- Gesellschaft, Philos., Erzieh.
- Informatik
- Jura
- Maschinenwesen
- Materialien
- Mathematik
- Medizin
- Physik
- Psychologie
- Sprache, Literatur und Kultur
- Umwelt
- Verkehr
- Weiterbildung
- Willkommen
- Wirtschaft
