this repo is a brief overview of what my team (github.com/pohkangyu, github.com/leechongyan and melissa) did for ndsc. the svm implementation is done by kangyu, while the neural networks were implemented by chong yan and i.
we did it on our local machines for the initial testing/prototyping of the models. however, due to the large size of the data set (more than 5gb), we shifted to using google cloud compute for the final model.
more information can be found in the pdf, which details our methodology for the final submission.