Nonstationary Forecasting Models


You are a regional manager of the ABC World Bank, and each year the cost of maintaining ATMs increases and depletes resources due to the additional cash demands during the holiday season. In order to increase efficiency in replenishing the machines, your corporate bank manager has asked you to forecast the daily cash demands (amount) and replenishing schedule for the month of December for your region’s 3 ATMs based on a previous December’s transaction history.

Unit 4 IP Tasks (Objectives from Week 1–4):
Step 1: Review the covered material from weeks 1-4 and the Web resources.

Step 2: Download the ATM Transactions dataset containing 14,913 total transactions from 3 locations. 

Step 3. Using SAS Studio and the provided ATM Transactions dataset, model and forecast the seasonal cash demands and replenishing schedule for a region’s 3 ATMs using SAS/ETS time series procedures. Take into consideration expected behavior of nonstationary models, fitting, and trending (seasonality may not apply since only one month’s worth of transaction history is being analyzed). 

Note: The actual cash demand (forecasted amount) will need to be based on the withdrawal patterns; therefore, it will be necessary to filter on the transaction type. However, the reloading schedule will need to be based on the transaction patterns for all transactions, in order to avoid inconveniencing customers during peak usage periods.

Step 4: Paste your (working) SAS code into a text document and submit. Add sufficient comments (see example below comment syntax) in your code about the model, each step, and any additional related notes. Include the appropriate coding for output (i.e., print=estimates, output=residuals, etc.) if applicable.

 * Use this syntax when adding comments. 
 * Comment block begins with a slash and an asterisk (*). 
 * Each line begins with an asterisk (*).
 * The comment block ends with an asterisk (*) and a slash. 

Do you need high quality Custom Essay Writing Services?

Order now