Farhad Shokoohi (Mathematical Sciences) and colleagues A. Khalili, M. Asgharian, both of McGill University, and S. Lin (Ohio State University) co-authored an article titled "Sparse estimation in semi-parametric finite mixture of varying coefficient regression models". The paper introduces an unsupervised learning and feature selection methods for high-dimensional data analysis when the effects of features are dynamic (age-dependent) and different in all intangible sub-populations. As an application, the paper gives an analysis of Osteocalcin data where genetic variants on the same chromosome are considered jointly.