Firstly, fantastic report. The style of writing was insightful yet not too difficult to follow and is reminiscent of research papers in this field. The visualizations were relevant and conveyed a good overview of the features and the data. The modeling was sound and analysis on the results was good-- especially the indication that the data was being overfitted.
Some criticisms include the relevancy of the topic itself; perhaps a heuristic approach would be all that is necessary for solving this problem (though it is too late to change topics). Another thing I'd like to point out is that the dataset is comparatively small, and the model would likely greatly benefit from added data from other sources (if possible). Otherwise, great work!
Firstly, fantastic report. The style of writing was insightful yet not too difficult to follow and is reminiscent of research papers in this field. The visualizations were relevant and conveyed a good overview of the features and the data. The modeling was sound and analysis on the results was good-- especially the indication that the data was being overfitted.
Some criticisms include the relevancy of the topic itself; perhaps a heuristic approach would be all that is necessary for solving this problem (though it is too late to change topics). Another thing I'd like to point out is that the dataset is comparatively small, and the model would likely greatly benefit from added data from other sources (if possible). Otherwise, great work!