This interface defines a scoring function that assigns a score to each input sequence.
SequenceScores, which can be used to score a
Sequence, typically using some model assumptions. Important special cases of
DifferentiableSequenceScores, which can compute the gradient with respect to their parameter for a given input sequence, including logistic regression, and
StatisticalModels, which can compute a proper (i.e., normalized) likelihood over the input space of sequences. These are located in the sub-package
StatisticalModels can be further differentiated into
TrainableStatisticalModels, which can be learned from a single input
DifferentiableStatisticalModels, which define a proper likelihood but can also compute gradients like