Using Deep Learning for FACT Source Detection
| dc.contributor.author | Bieker, Jacob | |
| dc.date.accessioned | 2019-05-20T18:45:58Z | |
| dc.date.available | 2019-05-20T18:45:58Z | |
| dc.date.issued | 2018-06 | |
| dc.description | Submitted to the Undergraduate Library Research Award scholarship competition: (2019). 80 p. | en_US |
| dc.description.abstract | Cosmic rays bombard the Earth constantly, causing air showers that contain information about the original particle and potentially about that particle's source. Determining if an air shower is from a gamma-ray or a hadron is a difficult problem to solve. Current methods primarily use a machine learning technique called random forests to determine whether a given event is from a gamma-ray or hadron, as well as the initial energy and source position in the sky by using the image an air shower makes in a detector. Another type of machine learning algorithm called neural networks has been shown to work very well on tasks involving images, in some cases outperforming random forests. This project aims to improve three tasks: determining the particle's type, energy, and source location using data from the First G-APD Cherenkov Telescope (FACT). | en_US |
| dc.identifier.uri | https://hdl.handle.net/1794/24579 | |
| dc.language.iso | en | en_US |
| dc.publisher | University of Oregon | en_US |
| dc.rights | Creative Commons BY-NC-ND 4.0-US | en_US |
| dc.title | Using Deep Learning for FACT Source Detection | en_US |
| dc.type | Thesis / Dissertation | en_US |