千葉大学 ChibaUniversity / 融合理工学府・先進理化学専攻・物理学コース
Search for Ultra-high Energy Neutrinos from Askaryan Radio Array (ARA) by Template Method
• Classified astronomical signal by statistical-oriented Principal Component Analysis (PCA), after obtaining features from 2 billion amounts (∼200 TB) of radio-frequency data measured below the South Pole. • Implemented automation solutions for utilizing large CPU & GPU clusters by building Python & C++ packages to streamline data analysis workflows and enhance productivity which has led to wildly use by international collaborators. • Implemented physics techniques, such as the Fast Fourier Transform (FFT), Interferometry, and the Matched Filter, into the package for feature extraction. • Optimized the PCA based on Frequentist Statistics and Pseudo Experiment. • Analyzed & quantified results by calculating statistical significance, including systematic uncertainty, and Monte Carlo simulation. • Performed high-precision calibration for the radio-frequency antenna for an advanced research instrument. • Learned large database management, including optimization of data sourcing and efficient connection to supercomputer by using solid analysis pipeline. • Practiced a thorough way to evaluate the project results by using statistical techniques and the back-of-the-envelope calculation.