Lattice-Based Minimum Error Rate Training Using Weighted Finite-State Transducers with Tropical Polynomial Weights
Aurelien Waite, Graeme Blackwood, Bill Byrne · 2012
Minimum Error Rate Training (MERT) is a method for training the parameters of a log-linear model. One advantage of this method of training is that it can use the large number of hypotheses encoded in a translation lattice as training data. We demonstrate that the MERT line optimisation can be modelled as computing the shortest distance in a weighted finite-state transducer using a tropical polynomial semiring. Finally we discuss the extension of our model to optimisation over multiple directions using an algorithm from Tropical Geometry. 1 1