Retrieved from https://studentshare.org/statistics/1694210-simple-moving-average-for-product-8
https://studentshare.org/statistics/1694210-simple-moving-average-for-product-8.
SIMPLE MOVING AVERAGE FOR PRODUCT 8 Utilizing SMA for weekly forecasting is simple to compute since an individual only require to divide the average price by the underlying period. The technique also offers a smoothed line free from whipsawing fluctuation in response to the prevailing tentative price dynamic. Moreover, it provides the sustainable level that depicts resistance. Product 8 records relatively higher sales within the 8d1 while the worst performing stores were 8a1 and 8c1. Running the test of the product with the data entail segregating the product classes into 8a1, 8b1, 8c1, 8d1and 8e11.
The strategy of the product is mainly based on the product name, store, product store id, sales, promotion, date, three weekly and corresponding four weekly. Moreover, selection of product is based on the long time duration. The forecasting accuracy of the technique relies on the valuable information of the product classes. Step 1 entails observing the 3 and 4 weekly sales volume of the products and make the decision on whether to subject the pattern to the pattern of seasonal alteration or selected necessary historical data2.
Step 2: smooth the sales trend with moving average by 3 week and taking the average and 4 week. Moreover, the moving average is performed to get the precise centered moving average of the week 3 and 4. Step 3 entails dividing the sales by the cma in order to attain the ratio that reflect the seasonal influence factor. For instance, the annual ratios of 8a product for 3 week are 7.66 and 4 week is 6.0. Product 8b had the forecasting accuracy of 13.67 and 15.5 while that of 8c1 is 4.67 and 3.5 correspondingly.
Moreover, product 8d1 had prediction accuracy of 20.33 and 22.5 while 8e1 was 4.667 and 5.75 respectively.ReferencesKlinker F, 'Exponential Moving Average Versus Moving Exponential Average' (2010) 58 Mathematische SemesterberichteTong H, 'Fitting A Smooth Moving Average To Noisy Data (Corresp.)' (1976) 22 IEEE Transactions on Information Theory
Read More