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:

PredictorProblem classDimensions
AffineLinear$M \rightarrow N$
AffineCombinationLinear$M \rightarrow N$
BinaryDecisionTreeMixed-integer linear$M \rightarrow 1$
GELUGlobal nonlinear$M \rightarrow M$
GrayBoxLocal nonlinear$M \rightarrow N$
Pipeline$M \rightarrow N$
QuantileLocal nonlinear$M \rightarrow N$
ReLUGlobal nonlinear$M \rightarrow M$
ReLUBigMMixed-integer linear$M \rightarrow M$
ReLUQuadraticNon-convex quadratic$M \rightarrow M$
ReLUSOS1Mixed-integer linear$M \rightarrow M$
ScaleLinear$M \rightarrow M$
SigmoidGlobal nonlinear$M \rightarrow M$
SoftMaxGlobal nonlinear$M \rightarrow M$
SoftPlusGlobal nonlinear$M \rightarrow M$
TanhGlobal nonlinear$M \rightarrow M$