Enviar Registro por Correo electrónico: Sparse generalized linear model with L(0) approximation for feature selection and prediction with big omics data