Tansu Daylan received his PhD in physics from Harvard University in 2018. He is a Kavli Fellow in Astrophysics at MIT.
Tansu’s main interest is in statistics with broad applications in cosmology and exoplanet research. During his PhD, he has worked on implementing “probabilistic cataloging”, a transdimensional and Bayesian inference framework that takes fair samples from the posterior probability distribution over the “element catalog space” given some observed data. When used to infer light sources consistent with observed photometric images, probabilistic cataloging can reveal faint and blended stars and galaxies, and perform robust error propagation by marginalizing over the uncertainties due to neighboring (covariant) sources.
He is also interested in the application of probabilistic cataloging to the gamma-ray sky, deep X-ray images and optical images of strong lens systems. Gamma-rays from the inner Milky Way yield a promising target for indirect searches for Weakly Interacting Massive Particles, i.e., a particle candidate for dark matter in the Universe. Images of strong lens systems, on the other hand, are informative about the spatial morphology of dark matter in self-gravitating halos.
At MIT, he has joined the TESS team to introduce probabilistic cataloging to images taken by TESS, enhance its data analysis pipeline to characterize systematic uncertainties and study planet candidates using the resulting light curves. He is also interested in inference using neural networks as applied to time series data, for classifying and characterizing features in light curves.