Vilalta, Ricardo2013-07-162013-07-16May 20132013-05http://hdl.handle.net/10657/405Space 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 kinds of research. There is considerable amount of data available; however, 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).application/pdfengThe author of this work is the copyright owner. UH Libraries and the Texas Digital Library have their permission to store and provide access to this work. Further transmission, reproduction, or presentation of this work is prohibited except with permission of the author(s).Pattern recognitionSpace PlasmaFASTSpace WeatherComputer science2013-07-16Thesisborn digital