In Partial Fulfillment of the Requirements for the Degree of
Master of Science
Will defend his thesis
Space plasma is made of electrically charged gases or fluids in space that are made up of free electrons and ions. They are studied extensively not only to analyze the dynamic processes of stellar bodies but also to understand various phenomena including particle acceleration, wave-particle interaction, applied science of space weather and its impact on human technology. The identification of primary particles of plasma is of utmost importance for these kind of research. There is considerable amount of data available, deriving a formula or methods for manual plasma regime identification is extremely time consuming, and can be highly unreliable and lack robustness. An automatic process of classifying these primary particles is of high demand. Currently, existing techniques that use machine learning algorithms have difficulty in distinguishing perceptible boundaries and regions as good as the human eye. In contrast, we propose a classification method to identify plasma particles automatically given a highly diversified time series data, based on energy and pitch angle. We came up with this algorithm after exploiting various learning techniques on the entire available data. Experiments are reported on datasets obtained from the Fast Auroral SnapshoT (FAST) explorer, which is the second mission in NASA’s Small Explorer Satellite Program (SMEX).
Date: Monday, April 15, 2013
Time: 2:00 PM
Faculty, students, and the general public are invited.
Advisor: Dr. Ricardo Vilalta