Non-parametric modeling of data

Ahmad Mehrabi

Department of Physics, Bu-Ali Sina University

Abstract: Given a data set, one can assume a model with some free parameters and then use a parameter inference method to find the best values of parameters as well as their uncertainties.

The maximum likelihood estimator and Bayesian inference provide two important methods to infer free parameters. Fixing all parameters of a model might result in losing some features or predicting biased values in those parts in which the data set is poor. On the other hand, one might develop some non-parametric modeling to reconstruct all possible curves that are consistent with the data. In this talk, I will review some non-parametric modeling of data and then apply them to Hubble data to reconstructing the Hubble diagram. Finally, I will compare these methods and give some discussions. 

یکشنبه 5 بهمن 1399، ساعت 15:00

Sunday 24 January 2021 – 15:00 Tehran Time

اتاق سمینار مجازی –Virtual Seminar Room

https://vc.sharif.edu/ch/cosmology

گزینه ورود به صورت مهمان – Enter as a Guest