UCB-Module-11-Assignment

Link to notebook

Overview

This repo contains data analysis on a Kaggle dataset about used car pricing. The goal of this analysis is to identify the key predictors for used car prices. And aims to provide a model for businesses to leverage in stocking vehicles and how to price them.

Dataset

• Source: Kaggle
• Description: The dataset contains information about used car prices, namely, categorical features of the car (make, model, year, etc.)
• Key Variables:
  • region: General region of the car's location (e.g. Prescott, Florida Keys)
  • price: Price of the car
  • year: Year of production of the car
  • manufacturer: Car manufacturer (e.g. Toyota, Ford)
  • model: Car model (e.g. Prius, Camry)
  • condition: Condition (e.g. Excellent, Fair)
  • cylinders: Number of engine cylinders (e.g. 6 cylinders, 8 cylinders)
  • fuel: Fuel type (e.g. gas, diesel)
  • odometer: Odometer reading (e.g. 12000)
  • title_status: Car title (e.g. clean, salvage)
  • transmission: Transmission type (e.g. automatic, manual)
  • VIN: VIN of the car
  • drive: Drive type (e.g. 4WD, AWD)
  • size: Car size (e.g. sedan, truck)
  • type: Vehicle type (e.g. car, truck)
  • paint_color: Color of the car.
  • state: State the car is located (e.g. California, Nevada).

Findings

• Age is the most important factor: The age of the car has by far the biggest impact on its price.
• Odometer reading is the second most important: The distance a car has been driven is also very influential.
• Drivetrain type matters: The type of drive (e.g. 2WD, 4WD) is a significant factor.
• Manufacturer plays a role: The car's manufacturer has notable importance.
• Engine and Vehicle Characteristics are Important: Features like the number of cylinders, model, fuel type, and condition also contribute substantially to price.
• Other Factors: Transmission type, condition, title status, and paint color have moderate importance.

Actionable Items

Pricing Strategy:

• Refine Age/Mileage Models: Non-linear depreciation, mileage bands, regional adjustments.
• Drive Type Pricing: Premium/discount based on local demand (2WD vs AWD/4WD).
• Manufacturer Adjustments: Value differences based on brand reputation.
• Dynamic Pricing: Real-time adjustments using market data.

Inventory Acquisition:

• Prioritize Lower Mileage: If cost-effective, target vehicles with lower odometer readings.
• Stock Popular Brands: Focus on manufacturers with strong resale value.
• Strategic Condition Sourcing: Balance high-end (excellent) with reconditionable vehicles.

Marketing & Sales:

• Highlight Age/Mileage: Emphasize low mileage, newer age vehicles.
• Targeted Ads: Segment by vehicle characteristics (AWD SUVs, etc.).

Ongoing Data & Analytics:

• Model Refinement: Regularly update pricing with new data.
• Competitor Analysis: Track similar vehicle prices.
• Demand Forecasting: Predict demand based on seasonality/economy.