On the mobile network running quality evaluation using UDM and Nearest Neighbo-Clustering RBFNN

This paper adopt Uniform Design Method (UDM) to design uniformity, representative and large-scale samples to train the RBF neural network (RBFNN) which is applied to carry out the mobile network quality evaluation. The experiments show, the difference between the real output of RBFNN evaluation and the expected output of Experts Fussy Evaluation of mobile network quality evaluation is very small. The designed evaluation method overcomes the difficulties of lack of large-scale and high quality training samples which causing the traditional RBF neural network´s poor generalization ability, and it realizes the non-linear approaching ability. meantime avoids the non-repeatability and instability of traditional experts evaluation.