Hosted by: Xin Qian

Data unfolding is a commonly used technique in the high energy physics experiments, to retrieve the distorted or transformed measurements by various detector effects. Inspired by the deconvolution technique in the digital signal processing, a new unfolding technique based on the Singular Value Decomposition (SVD) of the response matrix is developed. With the well-known Wiener filter concept, the modified SVD approach, Wiener-SVD, achieves the maximizing signal-to-noise ratio of the binned data in a transformed set of orthonormal bases where the uncertainties are bin-to-bin uncorrelated. In this talk, the mathematical principles and formulations of the newly developed Wiener-SVD unfolding will be presented. A few applications will be demonstrated. A comparison with the commonly used regularization method will also be shown. The advantages and disadvantaged of the Wiener-SVD approach will be discussed.