Application Of The Multi Index Comprehensive Evaluation Model Based On Factor Analysis And BP Neural Network

Comprehensive evaluation has been widely used in various fields of social e-conomy. After researching in many years, it gradually formed a variety of evaluation methods. Each method has its own advantages and disadvantages, therefore, it has been in the process of improvement and perfection continu-ously. In recent years, with the development and wide application of network, the information of data is increasing, which puts forward new requirements for the development of comprehensive evaluation methods. So the research on the multi index comprehensive evaluation method has more practical significance. The artificial neural network method is more popular these years. It has be-come a powerful tool to solve the complex and difficult modeling problem with its unique parallel distributed processing capabilities, nonlinear process-ing capabilities, self-learning function and fault tolerance. Base on the data’s several training, the output results directly determine the levels or categories for variables. It makes up the lack of relying on subjective judgment or ob-jective specific data for many previous methods to a certain degree, and it has become a widely used comprehensive evaluation method. When there are overmuch evaluation indexes, in order to improve the learning ability of the network, and speed up the processing and increase the reliability of evaluation, the network model needs to be simplified. When there are too many evaluation indexes, it can be matched with the statistical method effectively to reduce the dimensionality of the indexes, which can narrow down the outcome bias of neural network model in dealing with multi indexes and make the evaluation results more reliable.Combining factor analysis with BP neural network is to the improvement of BP neural network comprehensive evaluation method, which is the gist of this paper. The concrete thoughts is using factor analysis method for multi index dimension, then using the BP neural network model for comprehensive eval-uation. The model is applied to the evaluation of environmental air quality. AQI is the new standard of environmental air quality evaluation in2012. Ac-cording to AQI, by testing, it verifies the validity and feasibility of the multi index comprehensive evaluation model based on factor analysis and BP neural network. At the same time, it further verifies the practicability of the method by comparing with the neural network model. Moreover, the model can be further extended to multi index comprehensive evaluation in other areas.