Fedot parameters

[valer1435]

Please explain all parameters of Fedot classs @RepoPilotAssistant

[RepoPilotAssistant]

The Fedot class for FEDOT AutoML API has several parameters:

• problem: Specifies the type of modelling problem to solve. Options include:

  • classification for classification tasks
  • regression for regression tasks
  • ts_forecasting for time series forecasting tasks

• timeout: Defines the time limit for model design in minutes. None or -1 indicates no time limit.

• task_params: Additional parameters specific to the task.

• seed: Sets a fixed random seed for reproducibility.

• logging_level: Sets the logging level for the API. Options range from 50 (critical) to 0 (nonset).

• safe_mode: When set to True, it reduces dataset size to prevent memory overflow and uses label encoder instead of OneHot encoder for high cardinality categorical features. Default is False.

• n_jobs: Specifies the number of CPU cores to use for parallelization. -1 uses all available CPUs. Defaults to -1.

• composer_tuner_params: Additional optional parameters. Refer to the documentation for details at the methods of :class:~fedot.api.builder.FedotBuilder.