This project aims to simplify the process of pricing cars in Kenya for dealers and individual sellers.
Click on the stage to go to file.
| Step | Description | Tags |
|---|---|---|
| Project Formulation | This includes stating the problem I am solving that is pricing of cars in a competitive manner to attract buyers and make business sense of the market. Also included is the steps to be taken during the project undertaking |   |
| Data Sourcing | This involves scraping cheki.com website for car listings |    |
| Data Cleaning | From the scraped data the infromation we need is not readily available so we need to extract that information and come up with useful variables |   |
| EDA | Explored a number of features available in the car dataset and their relationship with price |      |
| Machine learning | Here I have explored Tree based and Linear Regression models and compared their fit on the data :Random forest has outperfromed other models by a significant margin in fitting the data :: From the results , a car's model eg whether it is a toyota,honda or BMW plays a significant role in pricing, the year of manufacture too is a significant price determinant among others. The r squared score of the final tuned model is an impressive .89 which means a significant portion of the variability in the car prices is explained by the model |    |
| Communication | to do |    |