DS-MLR: Exploiting Double Separability for Scaling up Distributed Multinomial Logistic Regression

Open in new window