... cross -domain language model adaptation paradigm, where we adapt a model trained on one domain (which we call the 228background domain) to a different domain (adap-tation domain) , for which ... of lasso for statistical language modeling for text input. Owing to the very large number of parameters, directly optimizing the pe-nalized lasso loss function is impossible. Therefore, we ... dependency and predictive clustering for language modeling. In EMNLP 2002. Gao. J., Yu, H., Yuan, W., and Xu, P. 2005. Minimum sample risk methods for language modeling. In HLT/EMNLP 2005....