To run the code, first put your buoy files in the input directory (begins with sp-)
afterwards, run python training_spark.py
an input file will be generated inside the feature_files directory
run python predict.py filename
where filename is the name of the file in the input directory (begins with sp-)