修正的广义学习向量量化算法

Revised GLVQ Algorithm

讨论了Pal等的广义学习量化算法（GLVQ）和Karayiannis等的模糊学习量化算法（FGLVQ）的优缺点，提出了修正广义学习量化(RGLVQ)算法。该算法的迭代系数有很好的上下界，解决了GLVQ的“Scale”问题，又不像FGLVQ算法对初始学习率敏感。用IRIS数据集对算法进行了测试，并应用所给算法进行了用于图像压缩的量化码书设计。该文算法与FGLVQ类算法性能相当，但少了大量浮点除法，实验过程表明节约训练时间约10%。

The advantage and defect of the generalized learning vector quantization (GLVQ) and fuzzy generalization learning vector quantization (FGLVQ) algorithms are discussed. A revised GLVQ (RGLVQ) algorithm is proposed. Because the interactive coefficients of the algorithms are properly bounded, the performance of the algorithms is invariant under uniform scaling of the entire data set unlike Pal's GLVQ, and the initial learning rate is not sensitive to the number of prototypes as Karayiannis's FGLVQ. The algorithm is tested and evaluated using the IRIS data set. The efficiency of the algorithm is also illustrated by its use in codebook design required for image compression based on vector quantization. The training time of RGLVQ algorithm is reduced by 10% as compared with Karayiannis's FGLVQ but the performance is similar.