Undergrad Research Project - Fall '14 Research

Fall 2014

Sean Bittner
Jelena Kovačević
Project description

This summer I worked on improving neural network reconstruction methods on calcium fluorescence data, and learned a lot about graph signal processing along the way. This fall, I will be working on a paper submission to the International Symposium on Biomedical Imaging which is due in November. The paper will detail a method for improving the speed and accuracy of standard neural network reconstruction techniques through matrix completion and a local thresholding technique.

Additionally, I will be working with Siheng Chen, analyzing various data-sets on which we can use graph signal processing to extract meaningful results. Particular data-sets we are interesting in investigating are U.S. temperature data, Netflix movie/tv show ratings, and professional networks in academia and industry.

The graph signal processing field is young, but I believe it has a lot of potential. Network signals can be found in nature as well as technology, and I believe the age of big data is far from over. I'm excited with the opportunity to continue my research this fall semester.

Return to project list