Skip to main content
Ctrl+K
pytorch-forecasting  documentation - Home pytorch-forecasting  documentation - Home
  • Getting started
  • Tutorials
  • Data
  • Models
  • Metrics
  • FAQ
  • Installation
  • API
  • Release Notes
  • GitHub
  • GitHub
Ctrl+K
  • Getting started
  • Tutorials
  • Data
  • Models
  • Metrics
  • FAQ
  • Installation
  • API
  • Release Notes
  • GitHub
  • GitHub
Ctrl+K

Section Navigation

  • data
    • encoders
      • _clipped_log
      • _clipped_logit
      • _identity
      • _minus_one
      • _plus_one
      • _square
      • softplus_inv
      • EncoderNormalizer
      • Expm1Transform
      • GroupNormalizer
      • MinusOneTransform
      • MultiNormalizer
      • NaNLabelEncoder
      • ReLuTransform
      • SoftplusTransform
      • TorchNormalizer
      • TransformMixIn
    • examples
      • _get_data_by_filename
      • generate_ar_data
      • get_stallion_data
    • samplers
      • GroupedSampler
      • TimeSynchronizedBatchSampler
    • timeseries
      • _coerce_to_dict
      • _coerce_to_list
      • _find_end_indices
      • check_for_nonfinite
      • TimeSeriesDataSet
  • models
    • base_model
      • _concatenate_output
      • _torch_cat_na
      • AutoRegressiveBaseModel
      • AutoRegressiveBaseModelWithCovariates
      • BaseModel
      • BaseModelWithCovariates
      • PredictCallback
      • PredictTuple
      • Prediction
    • baseline
      • Baseline
    • deepar
      • _deepar
        • DeepAR
    • mlp
      • _decodermlp
        • DecoderMLP
      • submodules
        • FullyConnectedModule
    • nbeats
      • _nbeats
        • NBeats
      • sub_modules
        • linear
        • linspace
        • NBEATSBlock
        • NBEATSGenericBlock
        • NBEATSSeasonalBlock
        • NBEATSTrendBlock
    • nhits
      • _nhits
        • NHiTS
      • sub_modules
        • init_weights
        • IdentityBasis
        • NHiTS
        • NHiTSBlock
        • StaticFeaturesEncoder
    • nn
      • embeddings
        • MultiEmbedding
        • TimeDistributedEmbeddingBag
      • rnn
        • get_rnn
        • GRU
        • LSTM
        • RNN
    • rnn
      • _rnn
        • RecurrentNetwork
    • temporal_fusion_transformer
      • _tft
        • TemporalFusionTransformer
      • sub_modules
        • AddNorm
        • GateAddNorm
        • GatedLinearUnit
        • GatedResidualNetwork
        • InterpretableMultiHeadAttention
        • PositionalEncoder
        • ResampleNorm
        • ScaledDotProductAttention
        • TimeDistributed
        • TimeDistributedInterpolation
        • VariableSelectionNetwork
      • tuning
        • optimize_hyperparameters
    • tide
      • _tide
        • TiDEModel
      • sub_modules
        • _ResidualBlock
        • _TideModule
  • metrics
    • base_metrics
      • convert_torchmetric_to_pytorch_forecasting_metric
      • AggregationMetric
      • CompositeMetric
      • DistributionLoss
      • Metric
      • MultiHorizonMetric
      • MultiLoss
      • MultivariateDistributionLoss
      • TorchMetricWrapper
    • distributions
      • BetaDistributionLoss
      • ImplicitQuantileNetwork
      • ImplicitQuantileNetworkDistributionLoss
      • LogNormalDistributionLoss
      • MQF2DistributionLoss
      • MultivariateNormalDistributionLoss
      • NegativeBinomialDistributionLoss
      • NormalDistributionLoss
    • point
      • CrossEntropy
      • MAE
      • MAPE
      • MASE
      • PoissonLoss
      • RMSE
      • SMAPE
      • TweedieLoss
    • quantile
      • QuantileLoss
  • utils
    • _dependencies
      • _check_matplotlib
      • _get_installed_packages
      • _get_installed_packages_private
    • _maint
      • _show_versions
        • _get_deps_info
        • _get_sys_info
        • show_versions
    • _utils
      • apply_to_list
      • autocorrelation
      • concat_sequences
      • create_mask
      • detach
      • get_embedding_size
      • groupby_apply
      • integer_histogram
      • masked_op
      • move_to_device
      • next_fast_len
      • padded_stack
      • profile
      • repr_class
      • to_list
      • unpack_sequence
      • unsqueeze_like
      • InitialParameterRepresenterMixIn
      • OutputMixIn
      • TupleOutputMixIn
  • API
  • models
  • deepar

deepar#

DeepAR: Probabilistic forecasting with autoregressive recurrent networks.

Modules

_deepar

DeepAR: Probabilistic forecasting with autoregressive recurrent networks which is the one of the most popular forecasting algorithms and is often used as a baseline

previous

baseline

next

_deepar

This Page

  • Show Source

© Copyright 2020, Jan Beitner.

Created using Sphinx 8.1.3.

Built with the PyData Sphinx Theme 0.16.1.