Learning Algorithm Sample Clauses

Learning Algorithm. The classification model, f , is trained by minimizing the loss function shown in Equation 1. However, this loss function uses the agreement model g, that also needs to be trained. We can think of g as regularizing the training process of f . Perhaps most interestingly though, while the agreement model can play a crucial role in training the classification model, the classification model can also help train the agreement model. A key contribution of our work is exactly this interaction in the training processes of f and g. More specifically, we propose the following learning algorithm: 1. Train the agreement model g to convergence, using the limited amount of labeled data that is provided. We refer to the initial trained model as g0.