We describe and analyze a simple and effective stochastic sub-gradient descent algorithm for solving the optimization problem cast by Support Vector Machines (SVM). We prove that the number of iterations required to obtain a solution of accuracy ɛ is Õ(1/ɛ), where each iteration operates on a

, and multiclass classification. First, we apply ideas from online sub-gradient methods [21, 19] and sequential dual optimization algorithms [11, 16], which are inclusive of the fastest algorithms for training linear SVMs and often significantly faster than the cutting plane algorithm used by SVM struct (a popular

models, and multiclass classification. First, we apply ideas from online sub-gradient methods [21, 19] and se-quential dual optimization algorithms [11, 16], which are inclusive of the fastest algorithms for training linear SVMs and often significantly faster than the cutting plane algo-rithm used

The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggesstions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA, 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any oenalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS. a. REPORT Heterogeneous multi-metric learning for multi-sensor fusion 16. SECURITY CLASSIFICATION OF: In this paper, we propose a multiple-metric learning algorithm to learn jointly a set of optimal homogenous/heterogeneous metrics in order to fuse the data collected from multiple sensors for classification. The learned metrics have the potential to perform better than the conventional Euclidean metric for classification. Moreover, in the case of heterogenous sensors, the learned multiple metrics can be quite different, which are adapted to each type of sensor. By learning the multiple metrics jointly within a single unified optimization