calhoun.analysis.crf
Interface CRFInference

All Known Implementing Classes:

SemiMarkovViterbi, SemiMarkovViterbiNoCache, Viterbi

public interface CRFInference

an interface to inference algorithms for CRFs. Given a model, a set of feature weights, and a set of input data, the algorithm selects a sequence of hidden states. The Viterbi dynamic programming algorithm and its variants are usually used for this problem.

Nested Class Summary

static class

CRFInference.InferenceResult holder which contains the results of an inference run.

Method Summary

CRFInference.InferenceResult

predict(ModelManager mm, InputSequence<?> data, double[] weights) Return the labelling that maximizes the conditional probability P(y|x).

Method Detail

predict

CRFInference.InferenceResult predict(ModelManager mm,
                                     InputSequence<?> data,
                                     double[] weights)

Return the labelling that maximizes the conditional probability P(y|x).

Parameters:

mm - model to use for training

data - input sequence to label

weights - array of feature weights. Usually these will be derived from a training pass.

Returns:

inference result containing the hidden states which are predicted and the score outputs.