Kai Liu has recently had his paper, “Spherical Principal Component Analysis,” accepted into SIAM International Conference on Data Mining SDM‘19.

This work propose a novel PCA formulation by adding a constraint on the factors to unify the Euclidean distance and the angle distance. Because the objective and constraints are nonconvex, the optimization problem is difficult to solve in general. To tackle the optimization problem, we propose an proximal alternating linearized minimization method with provable global sequence convergence and at least sublinear convergence rate. Experiments on synthetic data and real-world datasets have validated the effectiveness of our proposed method and demonstrated its advantages over state-of-art competing methods.

This work will be presented this May in Calgary, Alberta, Canada. SIAM Data Mining is one of the top conferences in data mining.