This paper utilizes Ant-Miner – the first Ant Colony algorithm for discovering classification rules – in the field of web content mining, and shows that it is more effective th...
Kernel Canonical Correlation Analysis (KCCA) is a method of correlating linear relationship between two variables in a kernel defined feature space. A machine learning algorithm b...
We examine linear program (LP) approaches to boosting and demonstrate their efficient solution using LPBoost, a column generation based simplex method. We formulate the problem as...
Ayhan Demiriz, Kristin P. Bennett, John Shawe-Tayl...
This paper investigates the use of a one-class support vector machine algorithm to detect the onset of system anomalies, and trend output classification probabilities, as a way to ...
Abstract. Sequential quadratic programming (SQP) methods have proved highly effective for solving constrained optimization problems with smooth nonlinear functions in the objective...
Philip E. Gill, Walter Murray, Michael A. Saunders