At&t I
At&t I
INTRODUCTION
The focus of this paper is on business forecasting techniques. Throughout this reading,
I used 6 distinct forecasting models which are Linear Regression, Simple Exponential
Smoothing ( = 0,1; 0,3; 0,45; 0,6), Decomposition with SES ( = 0,1), Decomposition with
Linear Regression, Moving Average, and Weighted Moving Average. Comparisons
between the various models will be made by looking at each models retrospective and forecast
R-Square, Mean Absolute Percentage Error (MAPE), AIC, BIC, Sum of Squared Error (SSE).
Being that there is a sufficient amount of data, a 12-month holdout period was used to evaluate
forecast accuracy.
The data forecasted was of AT&Ts daily stock price measured in US dollar. The data
used for forecasts ranged from March 18, 2016 to March 17, 2017, although as mentioned, a
12-month holdout period was use for evaluative purposes.
FORECASTING PROCESS