A4 Vertaisarvioitu artikkeli konferenssijulkaisussa
Speedy Local Search for Semi-Supervised Regularized Least-Squares
Tekijät: Gieseke F, Kramer O, Airola A, Pahikkala T
Toimittaja: Joscha Bach, Stefan Edelkamp
Konferenssin vakiintunut nimi: 34th Annual German Conference on AI
Julkaisuvuosi: 2011
Journal: Lecture Notes in Computer Science
Kokoomateoksen nimi: KI 2011: Advances in Artificial Intelligence: 34th Annual German Conference on AI, Berlin, Germany, October 4-7,2011. Proceedings
Tietokannassa oleva lehden nimi: KI 2011: ADVANCES IN ARTIFICIAL INTELLIGENCE
Lehden akronyymi: LECT NOTES ARTIF INT
Sarjan nimi: Lecture Notes in Computer Science
Vuosikerta: 7006
Aloitussivu: 87
Lopetussivu: 98
Sivujen määrä: 2
ISBN: 978-3-642-24454-4
eISBN: 978-3-642-24455-1
ISSN: 0302-9743
DOI: https://doi.org/10.1007/978-3-642-24455-1_8
Rinnakkaistallenteen osoite: http://research.utu.fi/converis/portal/Publication/3023302
In real-world machine learning scenarios, labeled data is often rare while unlabeled data can be obtained easily. Semi-supervised approaches aim at improving the prediction performance by taking both the labeled as well as the unlabeled part of the data into account. In particular, semi-supervised support vector machines favor decision hyperplanes which lie in a "low-density area" induced by the unlabeled patterns (while still considering the labeled part of the data). The associated optimization problem, however, is of combinatorial nature and, hence, difficult to solve. In this work, we present an efficient implementation of a simple local search strategy that is based on matrix updates of the intermediate candidate solutions. Our experiments on both artificial and real-world data sets indicate that the approach can successfully incorporate unlabeled data in an efficient manner.
Ladattava julkaisu This is an electronic reprint of the original article. |  |
