The paper introduces a new function generating a Bregman distance and develop Bregman proximal gradient methods applied to various matrix factorization problems. The authors clearly outline the benefits of the approach and computationally demonstrate this.