View source on GitHub
|
Keras losses in TF-Ranking.
Classes
class ApproxMRRLoss: Computes approximate MRR loss between y_true and y_pred.
class ApproxNDCGLoss: Computes approximate NDCG loss between y_true and y_pred.
class ClickEMLoss: Computes click EM loss between y_true and y_pred.
class CoupledRankDistilLoss: Computes the Rank Distil loss between y_true and y_pred.
class DCGLambdaWeight: Keras serializable class for DCG.
class GumbelApproxNDCGLoss: Computes the Gumbel approximate NDCG loss between y_true and y_pred.
class LabelDiffLambdaWeight: Keras serializable class for LabelDiffLambdaWeight.
class ListMLELambdaWeight: LambdaWeight for ListMLE cost function.
class ListMLELoss: Computes ListMLE loss between y_true and y_pred.
class MeanSquaredLoss: Computes mean squared loss between y_true and y_pred.
class MixtureEMLoss: Computes mixture EM loss between y_true and y_pred.
class NDCGLambdaWeight: Keras serializable class for NDCG.
class NDCGLambdaWeightV2: Keras serializable class for NDCG LambdaWeight V2 for topn.
class OrdinalLoss: Computes the Ordinal loss between y_true and y_pred.
class PairwiseHingeLoss: Computes pairwise hinge loss between y_true and y_pred.
class PairwiseLogisticLoss: Computes pairwise logistic loss between y_true and y_pred.
class PairwiseMSELoss: Computes pairwise mean squared error loss between y_true and y_pred.
class PairwiseSoftZeroOneLoss: Computes pairwise soft zero-one loss between y_true and y_pred.
class PrecisionLambdaWeight: Keras serializable class for Precision.
class RankingLossKey: Ranking loss key strings.
class SigmoidCrossEntropyLoss: Computes the Sigmoid cross-entropy loss between y_true and y_pred.
class SoftmaxLoss: Computes Softmax cross-entropy loss between y_true and y_pred.
class UniqueSoftmaxLoss: Computes unique softmax cross-entropy loss between y_true and y_pred.
class YetiDCGLambdaWeight: Keras serializable class for YetiDCGLambdaWeight.
class YetiLogisticLoss: Computes Yeti logistic loss between y_true and y_pred.
Functions
get(...): Factory method to get a ranking loss class.