Melbourne House Price Prediction

This repo is the summary for the work of Melbourne house price prediction.

Report

Under report folder, contain source code for reporting.

• report/README.Rmd: R markdown to provide data insight of Melbourne house price data.

Further report detail under report folder.

Data Processing

Under process folder, contain source code for data processing.

• process/spilt_train_data.R: R Script to split training data at local train set and local test set under 85/15 ratio.

Further processing detail under process folder.

Data Modeling

Under model folder, contain source code of models to predict Melbourne house price:

• model/model_decision_tree.R: R Script of Decision Tree Model.

• model/model_random_forest.R: R Script of Random Forest Model.

• model/model_gbm.R: R Script of Gradient Boosting Model.

Further modeling detail under model folder.

Data Source

Data source from Monash University, access the data from this competition vitticeps in Kaggle.

14 Features in this dataset:

1. id: unique id for property.

2. suburb: suburb location of property.

3. result: S indicates property sold; SP - property sold prior; PI - property passed in; PN - sold prior not disclosed; SN - sold not disclosed; NB - no bid; VB - vendor bid; o res - other residential; w - withdrawn prior to auction.

4. rating

5. nbeds: number of bedrooms.

6. property_type: h = house, t = townhouse, u = unit/apartment.

7. day: day of the month of auction.

8. month: month of auction.

9. year: year of auction.

10. nvisits: How many people came to open houses.

11. ncars: Number of parking places.

12. nbaths: Number of bathrooms.

13. land_size: Size of the lot, in sq m, units will be 0.

14. house_size: Internal size of property in sq m.