Scalable training ofL1-regularized log-linear models
Proceedings of the 24th international conference on Machine learning - ICML ’07
Jianfeng Gao
StatSnowball