Handwritten digits are recognized using prototypes created by atraining algorithm based on the Euclidean distance. The subsequentclassification of a handwritten digit is based on criteria consideringthe Euclidean distance to the prototypes. A training set of 2361patterns is used to create the prototypes and a separate set of 1320patterns is used to test the proposed method. The system performance iscompared to two other known classification algorithms: a MLP (multilayerperceptron network), and SOM (self-organizing map) plus LVQ1 (a linearvector quantization algorithm). The proposed method reached arecognition rate of 93.5% when using the nearest-prototype criterion,and raised to 94.8% when using a nearest-prototype-voting criterion. Itcompared favorably with the MLP (91.8%) and SOM+LVQ1 (91.5%)
展开▼
