Teaching data science and Bayesian statistics for physical sciences

I will describe the development and teaching experiences of the new course in UC Bereley physics department covering Bayesian statistics and machine learning applications in physical sciences. The course covers a broad range of topics including Bayesian uncertainty quantification and hypothesis testing, supervised and unsupervised machine learning, optimization and sampling methods, regression and classification methods, Fourier analyses etc. I will present several of the juptyer based homeworks and projects, where the methods are applied to examples in physics and astronomy.