In this project, you are given a dataset of real-estate properties, and we ask you to develop code that predicts the list price of a property. You should have in mind that the price of each property may change a lot given its region, and your solution should take that into account (try some feature engineering). For the solution, we request that you build at least 2 models for price prediction, and after evaluating each model, you should compare them to choose the best one for the task (use the mean absolute percentage error for comparison).
- Understanding the dataset
- Filter the dataset, remove unused, inconsistent and outliers values.
- Visualizing the dataset
- Choose 4 models from different categories
- Evaluation of models
- Conclusion
Please use Jupiter Notebook (price_prediction_challenge.ipynb) with explanation and complete code. The file in the html version (price_prediction_challenge.html) can also be used to read the methodology, contains the graphics but can not execute the code.