Car-Price-Prediction

Car Price Prediction model created based on Regression algorithms created with reference to DataTalks.Club mlzoomcamp course created using Google Colab

Dataset: https://raw.githubusercontent.com/alexeygrigorev/mlbookcamp-code/master/chapter-02-car-price/data.csv

1. Data Preparation

Reading the data from csv file and making initial data cleaning process of changing all texts to lowercase and removal of special characters and replacing spaces with underscore

2. Exploratory Data Analysis

• Checking for null values and examining important features.
• Feature engineering - adding new column to get better prediction.
• Using log expression to filter out less important areas. Data visualization using histplot

3. Model

• Implementing Randomforest regressor without pipleine using numerical features
• Implementing Randomforest regressor without pipleine using feature engineering
• Implementing Pipeline with Randomforest regressor algorithm using both all features
• Implementing Pipeline with XGBRegressor algorithm using both all features
• Implementing XGBRegressor without pipleine using both all features
• Trying out Crossvalidation

Picking the best model based on Mean Absolute Error (MAE) and Mean Squared Error(MSE)

Predicting the price of the car using the final model with test data

Will improve the model performance better in future!!