The machine-learning models, data sources, and validation principles behind Qiymetim.az.
Valuation based on comparable real listings
Indicators derived from collected vehicle listings
Price, mileage, year, and comparables assessed together
R², RMSE, and MAE are shown only with a reproducible report
Vehicle age and total mileage are model inputs.
Make, model, and engine displacement are considered.
Body style and fuel type are categorical inputs.
Accident history, transmission, and equipment are considered.
Listing data is compared with relevant market examples.
Invalid and corrupted figures are excluded from market statistics.
Important notes such as accidents, paintwork, and service needs are extracted.
Listing specifications, images, and description are collected.
Parameters are processed to estimate a fair price range.
Relevant facts are extracted and added to the final decision card.