RemixAutoML (version 0.5.0)

StackedTimeSeriesEnsembleForecast: TimeSeriesEnsembleForecast

Description

TimeSeriesEnsembleForecast to generate forecasts and ensemble data

Usage

StackedTimeSeriesEnsembleForecast(
  TS_Models = c("arima", "tbats", "nnet"),
  ML_Methods = c("CatBoost", "XGBoost", "H2oGBM", "H2oDRF"),
  CalendarFeatures = TRUE,
  HolidayFeatures = TRUE,
  FourierFeatures = NULL,
  Path = "C:/Users/aantico/Documents/Package",
  TargetName = "Weekly_Sales",
  DateName = "Date",
  NTrees = 750,
  TaskType = "GPU",
  GridTune = FALSE,
  FCPeriods = 5,
  MaxNumberModels = 5
)

Arguments

TS_Models

Select which ts model forecasts to ensemble

ML_Methods

Select which models to build for the ensemble

CalendarFeatures

TRUE or FALSE

HolidayFeatures

TRUE or FALSE

FourierFeatures

Full set of fourier features for train and score

Path

The path to the folder where the ts forecasts are stored

TargetName

"Weekly_Sales"

DateName

"Date"

NTrees

Select the number of trees to utilize in ML models

TaskType

GPU or CPU

GridTune

Set to TRUE to grid tune the ML models

FCPeriods

Number of periods to forecast

MaxNumberModels

The number of models to try for each ML model

See Also

Other Time Series Helper: FinalBuildArfima(), FinalBuildArima(), FinalBuildETS(), FinalBuildNNET(), FinalBuildTBATS(), FinalBuildTSLM(), GenerateParameterGrids(), OptimizeArfima(), OptimizeArima(), OptimizeETS(), OptimizeNNET(), OptimizeTBATS(), OptimizeTSLM(), ParallelAutoARIMA(), ParallelAutoArfima(), ParallelAutoETS(), ParallelAutoNNET(), ParallelAutoTBATS(), ParallelAutoTSLM(), PredictArima(), RL_Performance(), Regular_Performance(), TimeSeriesDataPrepare(), WideTimeSeriesEnsembleForecast()