Function reference

Sktime

Sktime models

Prophetverse	Univariate Prophetverse forecaster with multiple likelihood options.
HierarchicalProphet	A Bayesian hierarchical time series forecasting model based on Meta’s Prophet.

Exogenous effects

Exogenous effects

LinearEffect	Represents a linear effect in a hierarchical prophet model.
LinearFourierSeasonality	Linear Fourier Seasonality effect.
LogEffect	Represents a log effect as effect = scale * log(rate * data + 1).
HillEffect	Represents a Hill effect in a time series model.
MichaelisMentenEffect	Represents a Michaelis-Menten effect for modeling saturation.
ChainedEffects	Chains multiple effects sequentially, applying them one after the other.
GeometricAdstockEffect	Represents a Geometric Adstock effect in a time series model.

MMM Likelihoods

MMM Likelihoods

LiftExperimentLikelihood	Wrap an effect and applies a normal likelihood to its output.
ExactLikelihood	Wrap an effect and applies a normal likelihood to its output.

Trends

Trends

PiecewiseLinearTrend	Piecewise Linear Trend model.
PiecewiseLogisticTrend	Piecewise logistic trend model.
FlatTrend	Flat trend model.

Target Likelihoods

Likelihoods for the target variable

MultivariateNormal	Base class for effects.
NormalTargetLikelihood
GammaTargetLikelihood
NegativeBinomialTargetLikelihood

Budget Optimization

Budget Optimization

BudgetOptimizer

Budget optimizer using scipy.optimize.minimize.

Budget Constraints

Budget Constraints

TotalBudgetConstraint	Shared budget constraint.
MinimumTargetResponse	Minimum target response constraint.

Objective Functions

Objective Functions

MinimizeBudget	Minimize budget constraint objective function.
MaximizeKPI	Maximize the KPI objective function.
MaximizeROI	Maximize return on investment (ROI) objective function.

Budget Parametrizations

Budget Parametrizations