Learning under differing training and test distributions [Elektronische Ressource] / von Steffen Bickel

110

pages

English

Documents

2009

Écrit par
Bickel , Hans Steffen

Publié par
universitat_potsdam

Lire

Le téléchargement nécessite un accès à la bibliothèque YouScribe Tout savoir sur nos offres

110

pages

English

Ebook

2009

Lire

Le téléchargement nécessite un accès à la bibliothèque YouScribe Tout savoir sur nos offres

Publié par

universitat_potsdam

Publié le

01 janvier 2009

Nombre de lectures

Langue

English

Institut fu¨r InformatikArbeitsgruppe Maschinelles LernenLearning under Diﬀering Trainingand Test DistributionsDissertationzur Erlangung des akademischen Grades“doctor rerum naturalium”(Dr. rer. nat.)in der Wissenschaftsdisziplin Informatikeingereicht an derMathematisch-Naturwissenschaftlichen Fakult¨atder Universit¨at PotsdamvonSteﬀen BickelPotsdam, den 22.07.2009 Published online at the Institutional Repository of the University of Potsdam: URL http://opus.kobv.de/ubp/volltexte/2009/3333/ URN urn:nbn:de:kobv:517-opus-33331 [http://nbn-resolving.org/urn:nbn:de:kobv:517-opus-33331] AbstractOneofthemainproblemsinmachinelearningistotrainapredictivemodelfromtrain-ing data and to make predictions on test data. Most predictive models are constructedundertheassumptionthatthetrainingdataisgovernedbytheexactsamedistributionwhich the model will later be exposed to. In practice, control over the data collectionprocess is often imperfect. A typical scenario is when labels are collected by question-naires and one does not have access to the test population. For example, parts of thetest population are underrepresented in the survey, out of reach, or do not return thequestionnaire. In many applications training data from the test distribution are scarcebecause they are diﬃcult to obtain or very expensive.

Voir