paddlets.models.forecasting.dl.scinet

class SCINetModel(in_chunk_len: int, out_chunk_len: int, skip_chunk_len: int = 0, sampling_stride: int = 1, loss_fn: ~typing.Callable[[...], ~paddle.Tensor] = <function mse_loss>, optimizer_fn: ~typing.Callable[[...], ~paddle.optimizer.optimizer.Optimizer] = <class 'paddle.optimizer.adam.Adam'>, optimizer_params: ~typing.Optional[~typing.Dict[str, ~typing.Any]] = None, eval_metrics: ~typing.Optional[~typing.List[str]] = None, callbacks: ~typing.Optional[~typing.List[~paddlets.models.common.callbacks.callbacks.Callback]] = None, batch_size: int = 8, max_epochs: int = 100, verbose: int = 1, patience: int = 10, seed: ~typing.Union[None, int] = None, num_stack: int = 1, num_level: int = 3, num_decoder_layer: int = 1, concat_len: int = 0, kernel_size: int = 5, dropout_rate: float = 0.5, num_group: int = 1, hidden_size: int = 1)[source]

Bases: PaddleBaseModelImpl

DownSampled Convolutional Interactive Network (SCINet) for time series forcasting. Refers to SCINet .

Parameters

  • in_chunk_len (int) – The size of the loopback window, i.e. the number of time steps feed to the model.

  • out_chunk_len (int) – The size of the forecasting horizon, i.e. the number of time steps output by the model.

  • skip_chunk_len (int) – Optional, the number of time steps between in_chunk and out_chunk for a single sample. The skip chunk is neither used as a feature (i.e. X) nor a label (i.e. Y) for a single sample. By default, it will NOT skip any time steps.

  • sampling_stride (int) – Time steps to stride over the i-th sample and (i+1)-th sample. More precisely, let t be the time index of target time series, t[i] be the start time of the i-th sample, t[i+1] be the start time of the (i+1)-th sample, thus sampling_stride is equal to t[i+1] - t[i].

  • loss_fn (Callable[..., paddle.Tensor]|None) – Loss function.

  • optimizer_fn (Callable[..., Optimizer]) – Optimizer algorithm.

  • optimizer_params (Dict[str, Any], optional) – Optimizer parameters.

  • eval_metrics (List[str]|List[Metric], optional) – Evaluation metrics of model.

  • callbacks (List[Callback], optional) – Customized callback functions.

  • batch_size (int) – Number of samples per batch.

  • max_epochs (int) – Max training epochs.

  • verbose (int) – Verbosity mode.

  • patience (int) – Number of epochs to wait for improvement before terminating.

  • seed (int|None) – Global random seed.

  • num_stack (int) – stack number in Stacked SCINet.

  • num_level (int) – scinet tree level.

  • num_decoder_layer (int) – decoder layer number.

  • concat_len (int) – length to concat per stack.

  • kernel_size (int) – kernel size for Conv1D layer.

  • dropout_rate (float) – dropout regularization parameter.

  • num_group (int) – group number for Conv1D layer groups parameter.

  • hidden_size (int) – The number of features in hidden state for SCINet Interactor module.