I've used Machine Learning to create a model that predicts house prices in Melbourne :).
Libraries used include:
In further details, I have imported ensemble.RandomForestRegressor, metrics.mean_absolute_error and model_selection.train_test_split from sklearn
To minimize Mean Absolute Error (MAE), I chose RandomForestRegressor over DecisionTreeRegressor with various max_leaf_nodes 5, 50, 500, 1000. The result appeared to produce a minimum MAE between 50 and 500. I chose 250 as my nodes.