SELLER-LEVEL VEHICLE LISTINGS DATASET

The Seller-Level Vehicle Listings Dataset contains aggregated information about the vehicle inventories associated with individual sellers or dealers. Each record corresponds to a single seller, identified by sp_id, and summarizes the listings observed for that seller. The dataset combines seller-level statistics derived from vehicle characteristics, such as price, mileage, age, engine specifications, fuel economy, and vehicle dimensions, with indicators describing the composition of the seller's inventory, including the proportion of new, cab, salvage, and theft-title vehicles. It also includes temporal information about listing activity and categorical features representing the most common vehicle characteristics in each seller's inventory. Additional variables describe the seller's franchise status and its new/used inventory profile. The prediction target is seller_rating_class, a discretized seller-rating category with values low, medium, and high.

- The dataset for the project will be made available on the web page of the course.
- Detailed guidelines for the project will be presented during the course and made available on the web page of the course.

FEATURE DESCRIPTIONS

sp_id
Unique identifier of the seller/dealer. Used to aggregate vehicle listings at seller level.

n_listings
Total number of vehicle listings associated with the seller.

price_median
Median advertised price of the vehicles listed by the seller.

savings_amount_median
Median advertised savings amount of vehicles listed by the seller.

daysonmarket_median
Median number of days that the seller's vehicles remained listed on the market.

mileage_median
Median mileage of the vehicles listed by the seller.

year_median
Median model year of the vehicles listed by the seller.

vehicle_age_median
Median age of the vehicles at the time they were listed, computed as listing year minus vehicle model year.

owner_count_median
Median number of previous owners of vehicles listed by the seller.

horsepower_median
Median horsepower of vehicles listed by the seller.

engine_displacement_median
Median engine displacement of vehicles listed by the seller.

city_fuel_economy_median
Median city fuel economy of vehicles listed by the seller.

highway_fuel_economy_median
Median highway fuel economy of vehicles listed by the seller.

back_legroom_median
Median rear-seat legroom of vehicles listed by the seller.

front_legroom_median
Median front-seat legroom of vehicles listed by the seller.

fuel_tank_volume_median
Median fuel-tank capacity of vehicles listed by the seller.

height_median
Median vehicle height in the seller's inventory.

length_median
Median vehicle length in the seller's inventory.

width_median
Median vehicle width in the seller's inventory.

pct_isCab
Proportion of the seller's listed vehicles previously used as taxis or cabs.

pct_is_new
Proportion of the seller's listed vehicles classified as new vehicles.

pct_salvage
Proportion of the seller's listed vehicles having salvage status.

pct_theft_title
Proportion of the seller's listed vehicles having a theft-related title.

n_listing_days
Number of distinct calendar days on which listings from the seller are observed in the dataset.

listing_span_days
Number of days between the earliest and latest listing observed for the seller in the dataset.

dominant_body_type
Most frequently represented vehicle body type in the seller's inventory.

dominant_fuel_type
Most frequently represented fuel type among vehicles listed by the seller.

dominant_transmission
Most frequently represented transmission type among vehicles listed by the seller.

dominant_wheel_system
Most frequently represented drivetrain or wheel-system configuration in the seller's inventory.

dominant_listing_color
Most frequently represented standardized exterior color in the seller's inventory.

franchise_status
Seller category derived from the proportion of listings associated with a franchise dealer. Values are franchise, independent, mixed, or unknown.

new_used_profile
Composition of the seller's inventory with respect to new and used vehicles. Values are new_only, used_only, mixed, or unknown.

seller_rating_class
Discretized seller-rating category derived from seller_rating, with values low, medium, and high.
