Where academic tradition
meets the exciting future

You are here: TUCS > PUBLICATIONS > Publication Search > Speedy Local Search for Semi-S...

Speedy Local Search for Semi-Supervised Regularized Least-Squares

Fabian Gieseke, Oliver Kramer, Antti Airola, Tapio Pahikkala, Speedy Local Search for Semi-Supervised Regularized Least-Squares. In: Joscha Bach, Stefan Edelkamp (Eds.), KI 2011: Advances in Artificial Intelligence, Lecture Notes in Computer Science 7006, 87–98, Springer, 2011.

http://dx.doi.org/10.1007/978-3-642-24455-1_8

Abstract:

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.

Files:

Full publication in PDF-format

BibTeX entry:

@INPROCEEDINGS{inpGiKrAiPa11a,
  title = {Speedy Local Search for Semi-Supervised Regularized Least-Squares},
  booktitle = {KI 2011: Advances in Artificial Intelligence},
  author = {Gieseke, Fabian and Kramer, Oliver and Airola, Antti and Pahikkala, Tapio},
  volume = {7006},
  series = {Lecture Notes in Computer Science},
  editor = {Bach, Joscha and Edelkamp, Stefan},
  publisher = {Springer},
  pages = {87–98},
  year = {2011},
  keywords = {Semi-Supervised Learning, Kernel Methods, Regularized Least-Squares, Evolutionary Algorithms},
}

Belongs to TUCS Research Unit(s): Algorithmics and Computational Intelligence Group (ACI), Turku BioNLP Group

Publication Forum rating of this publication: level 1

Edit publication