Local learning framework for handwritten character recognition |
| |
| Affiliation: | 1. Centre for Pattern Recognition and Machine Intelligence, Concordia University, Montreal Que., Canada H3G 1M8;2. Department of Computer Science, Concordia University, 1455 de Maisonneuve Blvd W., Montreal Que., Canada H3G 1M8;1. Tetra Pak Processing Systems, Lund, Sweden;2. Lund University, Food Technology, Engineering and Nutrition, Lund, Sweden;3. Kristianstad University, Food and Meal Science, Kristianstad, Sweden;1. Department of Computer Science & Software Engineering, Faculty of Engineering and Computer Science, Concordia University, Montreal, Quebec, Canada, H3G 1M8;2. Department of Computer Engineering, Faculty of Electrical and Computer Engineering, University of Birjand, P.O. Box: 615/97175, Birjand, Iran;1. VeCAD Research Laboratory, Faculty of Electrical Engineering, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia;2. Machine Learning Developer Group, Sightline Innovation, #202, 435 Ellice Ave, Winnipeg, MB, Canada R3B 1Y6 |
| |
| Abstract: | This paper proposes a general local learning framework to effectively alleviate the complexities of classifier design by means of “divide and conquer” principle and ensemble method. The learning framework consists of a quantization layer which uses generalized learning vector quantization (GLVQ) and an ensemble layer which uses multi-layer perceptrons (MLP). The proposed method is tested on public handwritten character data sets, which obtains a promising performance consistently. In contrast to other methods, the proposed method is especially suitable for a large-scale real-world classification problems although it is easily scaled to a small training set while preserving a good performance. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
