Predictors
The main entry point for embedding prediction models into JuMP is add_predictor.
All methods use the form y, formulation = MathOptAI.add_predictor(model, predictor, x) to add the relationship y = predictor(x) to model.
Supported predictors
The following predictors are supported. See their docstrings for details:
| Predictor | Problem class | Dimensions |
|---|---|---|
Affine | Linear | $M \rightarrow N$ |
AffineCombination | Linear | $M \rightarrow N$ |
BinaryDecisionTree | Mixed-integer linear | $M \rightarrow 1$ |
GELU | Global nonlinear | $M \rightarrow M$ |
GrayBox | Local nonlinear | $M \rightarrow N$ |
Pipeline | $M \rightarrow N$ | |
Quantile | Local nonlinear | $M \rightarrow N$ |
ReLU | Global nonlinear | $M \rightarrow M$ |
ReLUBigM | Mixed-integer linear | $M \rightarrow M$ |
ReLUQuadratic | Non-convex quadratic | $M \rightarrow M$ |
ReLUSOS1 | Mixed-integer linear | $M \rightarrow M$ |
Scale | Linear | $M \rightarrow M$ |
Sigmoid | Global nonlinear | $M \rightarrow M$ |
SoftMax | Global nonlinear | $M \rightarrow M$ |
SoftPlus | Global nonlinear | $M \rightarrow M$ |
Tanh | Global nonlinear | $M \rightarrow M$ |