Customer_rating_model a well developed model designed for predicting feedbacks (or ratings) from the customer by using some correlated features are shown:
- product_category (
object) - product_subcategory (
object) - brand (
object) - delivery_status (
object) - assembly_service_requested (
bool) - payment_method (
object) - order_id (
int64) - customer_id (
int64) - product_price (
float64) - shipping_cost (
float64) - assembly_cost (
float64) - total_amount (
float64) - delivery_window_days (
int64)
A perfect ml model to boost your sales on new product by taking reviews. Feed model with your e-store's past traffic activities and get reviews on new product you wish to launch soon.
##Limitations
###error rate error rate can vary approx 1 to 1.5
###categorial restrictions
product_category must be from these:
-Outdoor
-Living Room
-Office
-Kitchen
-Bedroom
-Dining Room
product_subcategory must be from these:
-Bar Cart
-Pantry Cabinet
-Garden Chair
-Kitchen Cabinet
-Office Chair
-Dining Chair
-Lounge Chair
-Desk
-Kitchen Island
-Outdoor Table
-Bookshelf
-Dining Table
-Umbrella
-Sofa
-Mattress
-China Cabinet
-Buffet
-Computer Table
-Ottoman
-TV Stand
-Dresser
-Patio Set
-Side Table
-Filing Cabinet
-Nightstand
-Coffee Table
-Bed Frame
-Armchair
-Chest of Drawers
-Wardrobe
-Bar Stool
brand must be from these:
-Overstock
-HomeGoods
-World Market
-CB2
-IKEA
-West Elm
-Pottery Barn
-Ashley Furniture
-Urban Outfitters
-Crate & Barrel
-Wayfair
-Target
delivery_status must be from these:
-Delivered
-Failed Delivery
-Pending
-Cancelled
-In Transit
-Rescheduled
payment_method must be from these:
-Credit Card
-Apple Pay
-Cash on Delivery
-Debit Card
-Google Pay
-Bank Transfer
-PayPal
###Assumptions Based on the data, we assume that output rating must be from 1 to 5 and a Real Number.
##How To Use ?? Go to model.py file and see quickly how to use this model