| dc.description.abstract |
The use of existing data to support activities in decision making is not enough just to rely on operational data only, we need a data analysis to explore the potential of existing information. One of the data mining techniques that can be done is prediction. Prediction is needed to determine when an event will occur or arise, so that appropriate action can be taken. Prediction is not always 100% right, but with the right method selection can make predictions with a small error rate. Monica Water Depot is a business engaged in Refill Drinking Water. To find out the level of consumer use of refill drinking water, the business owner must know some predictions of upcoming purchases so that the business owner can provide gifts that are in accordance with the predicted purchases. Predictions on the level of consumer use can help employers provide gifts that are in accordance with consumer purchases, by knowing the purchase in the next period, the business owner can find out what gifts will be given to customers. The results of the analysis using I test data, II test data, and III test data obtained by the weight of moving avergae recommended using a weight of 2 months because it has the smallest error value when compared to using 3 months weight. Predictive results can be a recommendation for improvements made by management and as a means of determining business strategies in the future.
Keywords : Prediction, Data Mining, Weighted Moving Average, Average Forecasting Error Rate
| en_US |
