Task 1: Determine Store Formats for Existing Stores
- What is the optimal number of store formats? How did you arrive at that number?
The optimal number of store formats is 3. I arrived at this by running the K-Centroids Diagnostic Tool. Higher median, smaller variation
- How many stores fall into each store format?
- Based on the results of the clustering model, what is one way that the clusters differ from one another?
Cluster 1: This cluster sells more Deli and Meat than the other two clusters
Cluster 2: This cluster sells more Dairy, Floral and Frozen food than the other two clusters
Cluster 3: This cluster hardly makes more sales other categories than the first two clusters but makes major sales in General Merchandise compared to the other two
- Please provide a map created in Tableau that shows the location of the existing stores, uses colour to show clusters, and size to show total sales. Make sure to include a legend! Feel free to simply copy and paste the map into the submission template.
Task 2: Determine the Store Format for New Stores
- What methodology did you use to predict the best store format for the new stores? Why did you choose that methodology?
Boosted Model. The Boosted Model has the highest F1 score and highest accuracy rate for clusters 1, 2 and 3, so this model is the best option to predict the clusters for the 10 new stores
- What are the three most important variables that help explain the relationship between demographic indicators and store formats? Please include a visualization.
Age0to9
HVal750KPlus
Age65Plus
- What format do each of the 10 new stores fall into? Please provide a data table.
Task 3: Predicting Produce Sales
- What type of ETS or ARIMA model did you use for each forecast? Use ETS(a,m,n) or ARIMA(ar, i, ma) notation. How did you come to that decision?
After using the TS compare tool, I compared the error measurement of both ETS and ARIMA. I decided ETS was a better model because it had lower forecast error measurements. It has a lower RMSE, MAPE and MASE
Error Terms:
The in-sample error measures give us a look at how well our model is able to predict future values.
ETS:
ARIMA:
- Please provide a table of your forecasts for existing and new stores. Also, provide visualization of your forecasts that includes historical data, existing stores forecasts, and new stores forecasts.
| Month | New Stores | Existing Stores |
| Jan-16 | 2,563,357.91 | 21,136,641.78 |
| Feb-16 | 2,483,924.73 | 20,507,039.12 |
| Mar-16 | 2,910,944.15 | 23,506,565.98 |
| Apr-16 | 2,764,881.87 | 22,208,405.76 |
| May-16 | 3,141,305.87 | 25,380,147.77 |
| Jun-16 | 3,195,054.20 | 25,966,799.47 |
| Jul-16 | 3,212,390.95 | 26,113,792.57 |
| Aug-16 | 2,852,385.77 | 22,899,285.77 |
| Sep-16 | 2,521,697.18 | 20,499,583.91 |
| Oct-16 | 2,466,750.89 | 19,971,242.82 |
| Nov-16 | 2,557,744.59 | 20,602,665.92 |
| Dec-16 | 2,530,510.81 | 21,073,222.08 |