[mathematical modeling] Matlab Application Practice Series (95) - application cases of time series prediction (with matlab code)

Preface

Time series prediction is actually a regression prediction method , It's a quantitative prediction , Use the past time series data for statistical analysis , Guess where things are going .

The time series prediction method arranges the historical data of the prediction target into a time series according to the time sequence , Analyze their trends over time , And establish a mathematical model to extrapolate the quantitative prediction method . This paper introduces two methods of time series prediction , Moving average forecasting method and exponential smoothing forecasting method , These two methods are applicable to long-term regular time series .

01 The example analysis

In our country 1996~2006 The actual investment in is shown in the table below , Use the time series prediction method to predict 2007 year 、2008 Annual investment . And known 2007 The annual real investment amount is 137239, Compare the predicted value with the actual value .

year	1995	1996	1997	1998
Investment amount	20019.3	22913.5