/used_car_price

Primary LanguageJupyter Notebook

Used Car Price Model

This repository contains the end-to-end implementation for predicting used car price in the USA. This project is in active mode, so I keep experimenting ideas in notebook (in notebooks folder). I got the dataset from here on kaggle dataset page https://www.kaggle.com/austinreese/craigslist-carstrucks-data. Here are the words from the author of the data creator. Craigslist is the world's largest collection of used vehicles for sale, yet it's very difficult to collect all of them in the same place. I built a scraper for a school project and expanded upon it later to create this dataset which includes every used vehicle entry within the United States on Craigslist.

I have taken the data collected from the author and tried to implement it end-to-end with MLOps.

These are the steps I have performed:

  1. Data Cleaning
  2. Exploratory Data Analysis
  3. Feature Engineering
  4. Feature Selection
  5. Model building
  6. Model deployment using Streamlit
  7. Setup CI/CD for training automation
  8. Data and model versioning using DVC