timothynn / HousePricePrediction

I am using Machine Learning to create a model that predicts house prices in Melbourne :)
GNU General Public License v3.0
1 stars 0 forks source link
data-science machine-learning pandas random-forest scikit-learn xgboost

HousePricePrediction

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

Which Tree Model to use?

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.