What Is Uplift Modeling?
Uplift modeling estimates how an intervention changes an outcome for different people rather than predicting the outcome alone.
Don't Just Read it – Explore It
Uplift modeling estimates how an intervention changes an outcome for different people rather than predicting the outcome alone.
Survival analysis models the time until an event while accounting for cases whose event time is only partially observed.
Causal inference estimates how outcomes would differ under interventions, not merely which variables are associated in observed data.
A counterfactual explanation describes a feasible change to an input that would produce a different model outcome.
Bayesian optimization selects promising experiments by balancing uncertain model predictions against the value of exploring new settings.
Probabilistic programming lets developers describe uncertain models and use general inference tools to estimate unknown quantities.
Linear discriminant analysis separates classes with linear combinations of features under assumptions about class distributions and covariance.
A Gaussian process defines a probability distribution over functions and produces predictions together with uncertainty estimates.
A kernel method measures relationships between examples through a similarity function associated with a richer feature space.
The kernel trick computes inner products for a rich feature space without explicitly constructing every transformed feature.
A conditional random field predicts related labels together by modeling how observations and neighboring output choices interact.
Expectation-maximization alternates between estimating hidden assignments and updating parameters to improve a likelihood objective.