Motivation
A method for reliably detecting changes in the light curves of quasars (QSO) in real time and quantitatively describing these changes to facilitate easy understanding.
Methodology
- Conducting data transformation using a physics-based formula on irregularly observed astronomical time series data.
- The probability of being normal at each point in time is calculated using a distribution estimation model that utilizes conditional probabilities on the transformed data.
- Detect outliers when data deviating from normal probability is input and define this as a changing state.
Contribution
- Understanding the state changes of QSOs and identifying their physical mechanisms
- Selectively identifying reliable targets for spectroscopic follow-up observations to improve follow-up efficiency
- Understanding and exploring supermassive black holes
