清华大学-生产计划与控制：PPC-L04-Forecasting-2.pdfVIP

Lecture 4Lecture 4 ForecastingForecasting The Process of Predicting FutureThe Process of Predicting Future (2)(2) Ye ChengYe Cheng Dept. of Industrial EngineeringDept. of Industrial Engineering Tsinghua UniversityTsinghua University 1 Trend-Based Method Regression analysis Forecasting methods for trend Holt’s method Ye Cheng, Dept. IE, Tsinghua University           Production Planning  Control           Lecture 4  Forecasting (2) 2 Regression Analysis (1/2) n paired data points: (x , y ), (x , y ),…… (x , y ,) 1 1 2 2 n n xi , observed value of X y , observed value of Y i Y , dependent variable XX , independentindependent variablevariable Assumption: linear relationship between X and Y ˆ Linear regression Y a =+ bX The sum of the squared distances between the regression line and the ˆ data points is minimized , the predicted value of Y Y Ye Cheng, Dept. IE, Tsinghua University           Production Planning  Control           Lecture 4  Forecasting (2) 3 Regression Analysis (2/2) Forecasting problem: D , D , ... , D are the values of the demand 1 2 n at times 1, 2, ... , n The optimal values of a and b a DD − bb ((n + 1)1) // 22 bb SS // SS xy xx