Time-Series Forecasting

Overview

The dataset is provided by one of the largest fast-food restaurant chains in the US. It includes (1) transaction information such as menu items that were purchased and quantities of each item; (2) ingredient lists for individual menu items; (3) metadata on restaurants, including location, and store type. The data observation window is from early March, 2015 to 06/15/2015 and includes transactional data from 2 stores in Berkeley, CA and 2 stores in New York, NY.

Your job is to forecast the daily demand for the next two weeks (from 06/16/2015 to 06/29/2015) to help the managers make inventory replenishment decisions. In particular, you need to forecast demand for one specific ingredient: lettuce, for each of the four individual restaurants. You are expected to use both Holt-Winters and ARIMA methods in generating the forecast. The deliverable of the coursework includes a written report (submit both rmd file, and pdf/html file), and the forecast results summarized in a csv file.

Data Description

Table 1: pos_ordersale

• MD5KEY_ORDERSALE: String, Unique identifier for the order in POS
• ChangeReceived: Decimal, Change received from the sale
• OrderNumber: Integer, Order number, unique to the store
• TaxInclusiveAmount: Decimal, Tax amount added to the price of a transaction
• posTaxAmount: Decimal, Tax value
• MealLocation: Integer, Values:0-Eat in, 1-ToGo
• TransactionId: Integer, Transaction ID number
• StoreNumber: String, Store number (Corresponds to ’STORE_NUMBER’ in store_restaurant)
• date: Datetime, Date of the transaction

Row description: Franchisee point of sale transaction

Table 2: menuitem

• MD5KEY_MENUITEM: String, Unique identifier for the purchased menu item
• MD5KEY_ORDERSALE: String, Unique identifier for the POS order (Corresponds to ’MD5KEY_ORDERSALE’ in pos_ordersale)
• StoreNumber: Integer, Store number (Corresponds to ’STORE_NUMBER’ in store_restaurant)
• date: Datetime, Date of the transaction
• TaxInclusiveAmount: Decimal, Tax amount added to the price of an item
• TaxAmount: Decimal, Tax value
• AdjustedPrice: Decimal, Price of the item after discount adjustment
• DiscountAmount: Decimal, Amount of discount
• Price: Decimal, Original price
• Quantity: Integer, Quantity of the purchased item
• PLU: Integer, Unique identifier for the recipe associated with this menuitem (Corresponds with ’PLU’ in menu_items)
• CategoryDescription: String, Item category description (hierarchy of 3 levels)
• DepartmentDescription: String, More detailed item description
• Description: String, Most detailed item description
• Id: Integer, Unique identifier for associated menu item in menu_items

Row description: Purchased menu item associated with a single transaction in pos_ordersale Remark: Purchased menu items in menuitem are matched to menu items in menu_items on menuitem. PLU and menuitem.Id and menu_items.PLU and menu_items.MenuItemId

Table 3: store_restaurant

• STORE_NUMBER: String, Unique identifier for the store
• STORE_ADDRESS1: String, First line of store address
• STORE_ADDRESS2: String, Second line of store address
• DISTRIBUTION_REGION: String, Store’s region
• STORE_STATE: String, Store’s state
• STORE_ZIP: String, Store’s zip code
• STORE_TYPE: String, Indicates type of store

Row description: A store (or restaurant) location

Table 4: menu_items

• MenuItemId: Integer, Unique identifier for the menu item
• MenuItemName: String, Abbreviated description of the menu item
• MenuItemDescription: String, Detailed description of the menu item
• PLU: String, Unique identifier for the recipe associated with this menu item
• RecipeId: Integer, Unique identifier for the recipe (Corresponds with ’RecipeId’ in recipes)

Row description: A menu item associated with a franchisee recipe Remark: Purchased menu items in menuitem are matched to menu items in menu_items on menuitem. PLU and menuitem.Id and menu_items.PLU and menu_items.MenuItemId

Table 5: recipes

• RecipeId: Integer, Unique identifier for the recipe
• RecipeName: String, Abbreviated name of the recipe
• RecipeDescription: String, Full name of recipe

Row description: A recipe for a menu item

Table 6: recipes_ingredient_assignments

• RecipeId: Integer, Unique identifier for the recipe (Corresponds with ’RecipeId’ in recipes)
• IngredientId: Integer, Unique identifier for the ingredient (Corresponds with ’IngredientId’ in ingredients)
• Quantity: Decimal, Quantity of specific ingredient used in that recipe

Row description: A single recipe

Table 7: recipe_sub_recipe_assignments

• RecipeId: Integer, Unique identifier for the recipe (Corresponds with ’RecipeId’ in recipes)
• SubRecipeID: Integer, Unique identifier for the sub-recipe (Corresponds with ’SubRecipeID’ in sub_recipes)
• Factor: Decimal, Quantity of specific SubRecipeID used in RecipeID

Row description: A single sub-recipe

Table 8: sub_recipes

• SubRecipeId: Integer, Unique identifier for the sub-recipe
• SubRecipeName: String, Name of sub-recipe
• SubRecipeDescription: String, Description of sub-recipe

Row description: Sub-recipe associated with a recipe

Table 9: sub_recipe_ingr_assignments

• SubRecipeId: Integer, Unique identifier for the sub-recipe (Corresponds with ’SubRecipeId’ in sub_recipes)
• IngredientId: Integer, Unique identifier for the ingredient (Corresponds with ’IngredientId’ in ingredients)
• Quantity: Decimal, Quantity of ingredient used in the sub-recipe

Row description: A single ingredient within a sub-recipe

Table 10: ingredients

• IngredientId: Integer, Unique identifier for the ingredient
• IngredientName: String, Ingredient full name
• IngredientShortDescription: String, Ingredient short description
• PortionUOMTypeId: Integer, Portion unit of measure (Corresponds with ’PortionUOMTypeId’ in portion_uom_types)

Row description: An ingredient used in a franchisee recipe

Table 11: portion_uom_types

• PortionUOMTypeId: Integer, Unique identifier for the unit of measurement
• PortionTypeDescription: String, Text description of unit of measurement

Row description: Portion (or unit of measurement) for an ingredient