Method for predicting medicine sales trends on basis of medical big data

Abstract

The invention discloses a method for predicting medicine sales trends on the basis of medical big data. The method includes steps of (1), acquiring stationary time series data on the basis of the medical big data; (2), building autoregressive moving average models on the basis of the medical big data; (3), verifying residual series of time series models to determine whether the residual series arewhite noise series checking models or not; (4), building long-term trend prediction models and then computing periodicity indexes of medicines; (5), carrying out estimation by the aid of least squareprocesses to obtain periodicity index prediction models; (6), computing weights of proportions of the long-term trend prediction models and the periodicity index prediction models by the aid of standard deviation and acquiring combined prediction models; (7), analyzing, comparing and presenting prediction results. The method has the advantages that ARIMA (autoregressive integrated moving average)models, periodicity index models and the combined prediction models are built by the aid of the medical sales big data, accordingly, non-linear time-varying time series data can be accurately predicted, and the method has functions of predicting medicine sales.

Description

technical field [0001] The invention relates to the field of information technology, in particular to a method for predicting drug sales trends based on medical big data. Background technique [0002] Forecasting method is one of the most valuable research problems in the field of product retailing, and has a very wide range of applications in social and economic development, market trend analysis, product sales, etc. Some commonly used forecasting methods, such as ARIMA, BP neural network, and gray-scale forecasting can achieve better forecasting results. However, drug sales data is actually nonlinear and time-varying time series data. It is difficult for a single forecasting method to predict drug sales. Therefore, constructing a combination method that can predict long-term trends and short-term trends of time series data has always been a research focus in the field of drug sales forecasting methods. At present, for the drug sales forecast in the big data system, there ...

AI Technical Summary

Problems solved by technology

Some commonly used forecasting methods, such as ARIMA, BP neural network, and gray-scale forecasting can achieve better forecasting results. However, drug sales data is actually nonlinear and time-varying time series data. It is difficult for a single forecasting method to predict drug sales. Therefore, constructing a combination method that can predict both long-term trends and short-term trends of time series data has always been a research focus in the field of drug sales forecasting methods

At present, for the drug sales forecast in the big data system, there is still a lack of forecasting methods that can reflect the characteristics of the drug sales cycle and time characteristics

Images