Benchmarks
This folder contains a couple basic benchmarks to provide examples of reading in the data and creating a submission.
Executing these benchmarks requires python along with the pandas package.
To run them,
- Download the data
- Modify the get_paths function in util.py to point to the data path and the submission output path on your system
- Run the benchmarks by executing the corresponding script (e.g.
python random_forest_benchmark.py) - Make a submission with the output file
The benchmarks are:
- mean_benchmark.py: predicts the mean sale price from the training set
- median_benchmark.py: predicts the median sale price from the training set
- random_forest_benchmark.py: converts the training set to continuous and categorical features and then trains a random forest