Telephone traffic prediction method for calculating weighted moving average weight on the basis of univariate regression

Abstract

The invention discloses a telephone traffic prediction method for calculating a weighted moving average weight on the basis of univariate regression. On the basis of objective historical telephone traffic data, the weights of call center telephone traffic from Monday to Sunday are respectively calculated, and a telephone traffic prediction model is established. The invention adopts telephone traffic data in the last seven days in the historical telephone traffic data to directly obtain a telephone traffic prediction value of a first day after the historical telephone traffic data, and the telephone traffic prediction value of a second day after the historical telephone traffic data is obtained and so on. In a way of progressing according to date, the telephone traffic prediction method can calculate the telephone traffic prediction value after the historical telephone traffic data day by day through the telephone traffic prediction model. Therefore, the telephone traffic weight in the established telephone traffic prediction model does not depend on human subjective factors, such as experts and the like, and the telephone traffic prediction method has an advantage of high prediction precision and has a practical meaning for telephone traffic prediction.

Description

technical field [0001] The invention relates to a method for predicting traffic volume based on univariate regression calculation of weighted moving average weight. Background technique [0002] Call center traffic forecasting methods can generally be divided into qualitative forecasting and quantitative forecasting. Qualitative forecasts are guesses made using experience and intuition, which are highly subjective. Quantitative forecasting refers to the use of statistical methods to establish statistical models and analyze historical statistical data to make predictions for the future. In the past, research on dialogue traffic mainly used methods such as neural networks and time series. For example, the neural network method has disadvantages such as a large amount of calculation and difficult to explain in business language. The weight setting of the general weighted moving average method depends on human subjective factors such as experts. Contents of the invention ...

AI Technical Summary

Problems solved by technology

For example, the neural network method has disadvantages such as large amount of calculation and difficult to explain in business language. The weight setting of the general weighted moving average method depends on human subjective factors such as experts.

Images