Abstract
Finding the largest linearly separable set of examples for a given Boolean function is a NP-hard problem, that is relevant to neural network learning algorithms and to several problems that can be formulated as the minimization of a set of inequalities. We propose in this work a new algorithm that is based on finding a unate subset of the input examples, with which then train a perceptron to find an approximation for the largest linearly separable subset. The results from the new algorithm are compared to those obtained by the application of the Pocket learning algorithm directly with the whole set of inputs, and show a clear improvement in the size of the linearly separable subset obtained, using a large set of benchmark functions. © Springer-Verlag Berlin Heidelberg 2007.
Cite
CITATION STYLE
Franco, L., Subirats, J. L., & Jerez, J. M. (2007). MaxSet: An algorithm for finding a good approximation for the largest linearly separable set. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4668 LNCS, pp. 648–656). Springer Verlag. https://doi.org/10.1007/978-3-540-74690-4_66
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
