General Information

Project Description

In 2007, a DIMACS/DyDAn research team concluded that differences in entropy can be detected in cases of disease outbreaks. The goal of this research project is to use these results to create a reliable warning system which will provide early detection of disease outbreaks. Currently, we are working to define pre-processing parameters, to ensure maximum sensitivity of data. By adjusting window size, step size, and a binning algorithm, we hope to minimize error rates and maximize the effectiveness of our detection algorithm.

Weekly Log

Week 1: Did background reading to get acquainted with my problem. Prepared first presentation.

Week 2: Did more background reading, familiarized myself with entropy.

Week 3: Came up with elementary binning system with static bins.

Week 4: Brainstormed for new binning strategies. Tried using standard deviations.

Week 5: Realized many issues with current binning methods, need something new.

Week 6: Trying to find explicit equation for how probability of a symbol depends on window size and step size. Gave final presentation.

Week 7: Continued to work on equations. Prepared recap of summer progress.

Presentations