A Repeating Fast Radio Burst Selected by Machine Learning
橋本哲也 Tetsuya Hashimoto (NCHU)
※Keywords:
Radio Astronomy, Transients, Galaxies
※Description:
Fast radio bursts (FRBs) are mysterious radio flashes with millisecond timescales, most of which emerge in the extragalactic Universe. There are apparently two different populations of FRBs, including repeating and one-off FRBs.
Repeating FRBs could originate from the repeating activities of progenitors such as pulsars and magnetars, while one-off FRBs may be generated by catastrophic one-off events such as supernovae and compact merger systems.
Therefore, a key element in revealing the origin of mysterious FRBs is to understand their repetitive nature. However, it requires expensive long monitoring, which is not feasible within the limited amount of telescope time. To address this issue, we proposed a new technique to select repeating FRB candidates from single-epoch FRB observation with machine learning. To prove our method, we conducted follow-up observations with Five-hundred-meter Aperture Spherical Telescope (FAST) for one of the repeating FRB candidates selected by machine learning. The student will work on this data to find FRBs. Once proven by FAST, our method will be extremely useful because the machine learning method does not require long monitoring.
If no FRB signal is found in the data, the student will switch to one of the statistical approaches proposed in another project plan. Following successful cases of past summer school students, we plan to publish a paper based on this project. The supervisor will provide the student with a working desk in his office to maximize the research efficiency during the program.
※Required Background:
Unix/Linux experience and programming skills such as Python are recommended, although learning them is a part of the summer student program.
Basic background in physics.