Project Name: Price Prediction of Used Cars
Project Members: Wei Cheng (wc655), Jiahe Chen (jc3472)
The question we want to explore from this data set is how to predict the price of used cars based on features like make, year, wheel drive, etc. Although such kind of price prediction can be found everywhere on used car websites, these websites never release their prediction methods and we do not know which features impact the price most. Answering this question can provide a more transparent guideline to customers and also help the car owners know which parts of the car should be maintained properly or when the car should be sold so that the owner can argue for a better price.
We use this dataset from Kaggle, which contains the most all relevant information that Craigslist provides on car sales including columns like price, condition, manufacturer, latitude/longitude, and 18 other categories. The dataset also has more than 400,000 data points. Therefore, it provides enough data for us to train, validate and test different models.