Lab 277

INTEROFFICE ME MORANDU M

TO:

PROF. MCCLINTOCK

FROM:

MEGAN LEE

SUBJECT:

CHAPTER 18 LAB

DATE:

DECEMBER 4, 2018

Introduction As requested, this document analyzes sales data from a major retailer in Brazil to accurately forecast demand for the product. This will enable the retailer to order appropriate levels of merchandise in stock. The exponential smoothing technique was used on this time series data and any specific quarterly effects on sales were also identified. Data Analysis

Using the exponential smoothing method comparison helps address the unsystematic portion of the data. A= 0.3 A = 0.8 MSE = 4074.6123 MSE = 3969.0072 MAD = 44.357681 MAD = 41.921419 In order to choose the best model, we can compare the two models and the we would want to choose the model with the smaller MSE and MAD, which means that we have less errors. Based on these two models, we should use the second model of A = 0.8.

Regression equation: ̂ = 66.2607 + 2.53892(𝑑1) + 0.66726(𝑑2) − 16.43461(𝑑3) 𝑆𝑎𝑙𝑒𝑠 R-squared: We are 3% of the way to perfectly forecasting demand using this model Standard-error: We are 73.26% off on our predictions using this model, on average all else constant. Q3 is significant.

2

Conclusion This data showed the exponential smoothing method to address the systematic and unsystematic patterns of the data. To analyze the systematic patterns of the data, the seasonal dummy variable forecasting method was used. From the data, Q3 was significant in forecasting average sales. This is helpful for companies to predict sales and adjust their inventory accordingly. If there are any further questions, please feel free to reach me at [email protected].

3