https://github.com/nirav2002/used_car_price_prediction_system
Created using various Machine Learning models on Jupyter notebook
https://github.com/nirav2002/used_car_price_prediction_system
machine-learning
Last synced: over 1 year ago
JSON representation
Created using various Machine Learning models on Jupyter notebook
- Host: GitHub
- URL: https://github.com/nirav2002/used_car_price_prediction_system
- Owner: nirav2002
- Created: 2022-12-23T10:21:05.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2023-04-12T12:35:50.000Z (about 3 years ago)
- Last Synced: 2025-02-03T22:47:58.019Z (over 1 year ago)
- Topics: machine-learning
- Language: Jupyter Notebook
- Homepage:
- Size: 4.95 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Used_Car_Price_Prediction_System
---
The process of selling a car at the right price is tedious and takes a toll on the general public. No one wants to sell their vehicle below its deserved price, and there is currently no model that helps the public know the selling price of their vehicles. We developed a model that will give customers the right price for their vehicles. The customer has to enter the car’s specifications, such as the sunroof system, 4-wheel drive, auto parking, and various other features. The customer can also decide how recent or old the car is. Now, with all these inputs, our system will help the customer by predicting the best possible price for their vehicle by implementing machine learning.
According to market research, the price of your vehicle will depreciate by 5% of the purchase price and may even depreciate by 10% later on, depending on a variety of factors. It's also said that the price of your car can go down to 50% of its original cost within 4 to 5 years. While all of these facts are correct, getting a good price for the customer becomes an even more difficult task. Most of the customers will sell their vehicle for a lower price than what it is capable of being sold for. The major reason is that the customer has no idea how to get information about the accuracy of the price offered, i.e., in other words, the customer will not know if the price offered is the right price or not. Hence, this led us to develop the used car price prediction model. This model will not only predict the accuracy of the price offered, but will also help the customers make a decision quicker and easier when it comes to selling their vehicle.
We have used the dataset from kaggle, whose link is provided in a seperate file inside the project. We have performed Data Pre-Processing on the dataset, to remove the occurence of null values so as to get better accuracy. We have also imputed values in certain columns of the dataset, as for those particular columns, we felt removing the null values could hamper the accuracy of the model. Outliers also have been removed in the Pre-Processing Phase.
We have splitted the dataset into testing (70%) and training (30%). After this, we have used various Machine Learning Models for prediction of the correct price of the vehicle. We have successfully calculated the Mean Squared Error, Mean Absolute Error, R-Squared Error and Root Mean Squared Error for Linear, Ridge, Lasso, Bayesian Regression, Decision Tree, Random Forest Regression, Muilti-layer Perceptron (MLP), Light gradient-boosting, XG Boost, Gradient Boosted Regression.
These different errors have been plotted to get a clear and a concise view of how the different algorithms work on our dataset.