Evaluation of Smart City Developmental Level Based on Principal Component Analysis and GA-BP Neural Network

Abstract

Smart city assessment issue is an important component of smart city construction. On one hand, it can help the government to guide and direct the activities of Smart-city Construction, on the other hand, it can reflect and give feedbacks to the audience. In this paper, according to the existing evaluation system of Smart City at home and abroad and the division standard of the latest cities in China, we create a more complete and comprehensive evaluation system. At first, we use the Principal Component Analysis (PCA) to reduce index that is according to design the evaluation index of smart city developmental level . Then, these index after reducing let input BP neural network optimized by Genetic Algorithm to train and simulate, find the error of between the actual output value and expected value reach the expected goal. At last, we use directly BP neural network and compare the errors and find using GA-BP neural network prefer. Thus further proves the scientificity and rationality of the evaluation method.