OCDocker.OCScore.DNN.future package

Submodules

• OCDocker.OCScore.DNN.future.DNNOptimizer module
  • DNNOptimizer
    • DNNOptimizer.__init__()
    • DNNOptimizer.mask
    • DNNOptimizer.rank_train
    • DNNOptimizer.rank_val
    • DNNOptimizer.from_embeddings()
    • DNNOptimizer.objective()
    • DNNOptimizer.optimize()
    • DNNOptimizer.set_random_seed()
• OCDocker.OCScore.DNN.future.datasets module
  • EnergyDataset
    • EnergyDataset.__getitem__()
    • EnergyDataset.__init__()
    • EnergyDataset.__len__()
  • TargetRankingDataset
    • TargetRankingDataset.__getitem__()
    • TargetRankingDataset.__init__()
    • TargetRankingDataset.__len__()
  • TargetBatchSampler
    • TargetBatchSampler.__init__()
    • TargetBatchSampler.__iter__()
    • TargetBatchSampler.__len__()
• OCDocker.OCScore.DNN.future.losses module
  • UncertaintyWeighting
    • UncertaintyWeighting.__init__()
    • UncertaintyWeighting.forward()
  • focal_binary_loss()
  • lambda_rank_ndcg_loss()
  • supervised_contrastive_loss()
• OCDocker.OCScore.DNN.future.metrics module
  • compute_classification_metrics()
  • compute_group_metrics()
  • ndcg_at_k()
  • partial_auc()
  • safe_auc()
  • safe_log_loss()
  • safe_pr_auc()
• OCDocker.OCScore.DNN.future.models module
  • MLP
    • MLP.__init__()
    • MLP.forward()
  • MultiTaskModel
    • MultiTaskModel.__init__()
    • MultiTaskModel.forward()
  • parse_encoder_params()

Module contents

Future DNN pipeline modules (multi-stage, multi-task).

Usage:

import OCDocker.OCScore.DNN.future as ocdnn_future

Modules

• DNNOptimizer: Experimental DNN optimizer.

• datasets: Dataset helpers for experiments.

• losses: Experimental loss functions.

• metrics: Experimental metrics.

• models: Experimental model definitions.