A neural support vector machine |
| |
Authors: | Magnus Jändel |
| |
Institution: | 1. Department of Radiation Oncology, UCLA Medical Center, Los Angeles, CA;2. Department of Interventional Radiology, UCLA Medical Center, Los Angeles, CA;1. Mathematics Department, Duke University, Box 90320, Durham, NC 27708-0320, United States;2. School of Mathematics, University of Minnesota, 127 Vincent Hall, 206 Church St. SE, Minneapolis, MN 55455, United States;1. School of Mathematics, University of Minnesota, 127 Vincent Hall, 206 Church St. SE, Minneapolis, MN 55455, USA;2. Mathematics Department, Duke University, Box 90320, Durham, NC 27708-0320, USA |
| |
Abstract: | Support vector machines are state-of-the-art pattern recognition algorithms that are well founded in optimization and generalization theory but not obviously applicable to the brain. This paper presents Bio-SVM, a biologically feasible support vector machine. An unstable associative memory oscillates between support vectors and interacts with a feed-forward classification pathway. Kernel neurons blend support vectors and sensory input. Downstream temporal integration generates the classification. Instant learning of surprising events and off-line tuning of support vector weights trains the system. Emotion-based learning, forgetting trivia, sleep and brain oscillations are phenomena that agree with the Bio-SVM model. A mapping to the olfactory system is suggested. |
| |
Keywords: | |
本文献已被 ScienceDirect 等数据库收录! |
|