Abstract:

Grey prediction is one of the commonly used algorithms for forecasting and analyzing situation data, but the accuracy of traditional grey prediction model is low in forecasting low smoothness data series. To solve this problem, a difference weighted smoothing method was proposed, and the Euler correction method was used to compensate for the error. The improved grey prediction model was tested on a library borrowing database. The test and analysis results showed that the variance of the relative error of the prediction results of the model had been reduced from 0.32 to 0.142, and the absolute value of the average relative error had been reduced by 4.458%. The improved grey prediction model could process the library borrowing data with low smoothness, predict it accurately, and provide a reference basis for improving the work efficiency and development and construction of the library.

Key words: grey prediction, weighted difference, smoothness, Euler correction, whitening differential equation