Vehicle Price Prediction Model This repository contains a complete Machine Learning pipeline to predict vehicle selling prices using the Vehicle Dataset from CarDekho. The project includes all necessary steps, from data loading and cleaning to model training and evaluation, following best practices in machine learning.
Project Overview The goal of this project is to build a model that accurately predicts the selling price of vehicles based on various features such as age, fuel type, transmission type, and more.
Steps Included: Data Download and Loading:
Downloading the dataset from Kaggle and loading it into a Jupyter Notebook for exploration. Data Preparation and Cleaning:
Handling missing values and dropping unnecessary columns to prepare the data for analysis. Encoding categorical features for compatibility with the ML model. Feature Engineering:
Modifying columns (e.g., calculating vehicle age from the year column) to enhance the dataset for better model training. Data Visualization:
Visualizing relationships and distributions in the data to understand patterns that may influence the target variable (selling_price). Data Splitting:
Splitting the dataset into training and test sets to evaluate model performance on unseen data. Model Training:
Using a Random Forest Regressor to predict the selling_price. Evaluating the model using metrics like Mean Absolute Error (MAE), Mean Squared Error (MSE), and R² Score on the training and test datasets. Model Evaluation and Scope for Improvement:
Assessing model performance on training data. Identifying potential ways to improve the model, such as experimenting with different algorithms, tuning hyperparameters, or adding feature engineering steps. Usage Clone the repository. Download the dataset from Kaggle, and add it to the project directory. Run the notebook to follow each step and see the results.