Unable to view the PDF below?
Try a direct download.

# 841 Paper
# 841 Score Report

 Dear Team #841,
 
You created a reasonable and thoughtful report to model the two possible scenarios mathematically. Although a bit wordy, your summary clearly and effectively stated the interpretation and solution for both problems. Your paper also included great model settings, comprehensive explanations of the methodology and solutions, and benefited from the detailed examples of computations and commentary. However, some typos need to be addressed in the report (e.g., some part of the math formula (3) is missing). You could have also included more examples to illustrate some of your steps, making it easier for the reader to understand your ideas better. Just as what you stated in your weakness section, you still need to explain more about your choice of the arctan function. Maybe test it using more data points.

 Another strength of this paper is that you calibrate the assumptions of the model using the data from previous elections. Although this method is most optimal, unfortunately, your explanations were not clear enough, forcing readers to look directly at your program. The explanation could be strengthened by including additional discussion on the sensitivity of your model to people's certainty factors. For example, how the candidate’s success rate will be affected if the voters’ certainty levels decrease from 50% to 45%?

For problem two, you need to explain more on how the last few steps are done in your summary. Fortunately, the model is explained in details with the help of some graphs, and the math symbols are clearly defined. I also liked that you provide the code and that there are calibrations with data. The assumptions and model are reasonable, but the paper would benefit from including a brief overview of the algorithms used and additional discussion regarding the sensitivity of the model. Some math formulas (and formatting) crashed when the file was converted to a PDF. Also, some background explanation of Dijkstra algorithm should be presented.

Introducing T seems unnecessary, as you can just take the number of accidents in the area and divide it by the total number of accidents in Manhattan. The list of results on page 21 also needs more explanation. Similar to problem one, the sensitivity analysis section can discuss the impact of having multiple accidents at the same time (in which case we will need to dispatch an ambulance from the next nearest hospital) on the expected transportation time. Your second strength claims that your paper considered traffic jam, but I could not find it anywhere in your model.

Some of your appendices are unnecessary, such as the assigned value to each edge. Commenting next to your code in English will help the readers better understand your approach.

Overall, I enjoyed reading your paper.
 
Best,
Association of Computational and Mathematical Modeling