Stochastic Gradient Descent Training for L1-regularized Log-linear Models with Cumulative Penalty.
Yoshimasa TsuruokaJun'ichi TsujiiSophia AnaniadouPublished in: ACL/IJCNLP (2009)
Keyphrases
- stochastic gradient descent
- least squares
- log linear models
- loss function
- matrix factorization
- step size
- regularization parameter
- objective function
- weight vector
- support vector machine
- multiple kernel learning
- feature selection
- latent variables
- conditional random fields
- mixture model
- principal component analysis
- probabilistic model
- discriminative training
- machine learning