Conclusion:
In this study, machine learning was used to predict the treatment an
individual patient will likely undergo using pre-treatment variables
obtained in the NCDB. In doing so, this study demonstrates how the use
of machine learning can be broadened in prediction modeling for future
analysis within clinical research. This study also highlights that the
largest influencers determining the treatment for an individual patient
are tumor characteristics with minor influencers including type of
facility that treatment was sought at.