MixedEffectsModel<T>

Models & Types · AiDotNet.Regression

Mixed-Effects (Hierarchical/Multilevel) Linear Model for clustered and hierarchical data.

Inheritance: NonLinearRegressionBase<T> → MixedEffectsModel<T>

For Beginners

Use this model when your data has groups or clusters:

Model structure: y_ij = X_ij * β + Z_ij * u_j + ε_ij

Where:

  • y_ij: Outcome for observation i in group j
  • X_ij * β: Fixed effects (same for everyone)
  • Z_ij * u_j: Random effects (vary by group)
  • ε_ij: Residual error

Example - Student test scores: score = study_hours * β + (school_intercept + study_hours * school_slope) + error

Fixed effects tell you: “On average, each study hour adds β points” Random effects tell you: “School 1 starts 5 points higher, School 2 has steeper study benefit”

Key outputs:

  • Fixed effect coefficients and their standard errors
  • Random effect predictions (BLUPs) for each group
  • Variance components showing how much groups vary

How It Works

Mixed-effects models combine fixed effects (population-level patterns) with random effects (group-level variations). They properly account for the correlation structure in hierarchical data, providing correct inference when observations are not independent.

Example

using AiDotNet;
using AiDotNet.Data.Loaders;
using AiDotNet.Regression;
using AiDotNet.Tensors.LinearAlgebra;

double[][] features =
{
    new[] { 1.0, 2.0 }, new[] { 2.0, 3.0 }, new[] { 3.0, 4.0 },
    new[] { 4.0, 5.0 }, new[] { 5.0, 6.0 }, new[] { 6.0, 7.0 }
};
double[] targets = { 3.0, 5.0, 7.0, 9.0, 11.0, 13.0 };

var result = await new AiModelBuilder<double, Matrix<double>, Vector<double>>()
    .ConfigureModel(new MixedEffectsModel<double>())
    .ConfigureDataLoader(DataLoaders.FromArrays(features, targets))
    .BuildAsync();

Console.WriteLine("Trained MixedEffectsModel.");

Constructors

Constructor Summary
MixedEffectsModel(MixedEffectsModelOptions options, IRegularization<T, Matrix<T>, Vector<T>> regularization) Initializes a new instance of Mixed-Effects Model.

Properties

Property Summary
T ResidualVariance { get; } Gets the residual variance.
bool SupportsParameterInitialization { get; } MixedEffects doesn’t support optimizer parameter injection.

Methods

Method Summary
Dictionary<int, Vector<T>> GetRandomEffects() Gets the random effects (BLUPs) for each group.
IEnumerable<int> GetActiveFeatureIndices() Returns all features used by the model.
Matrix<T> GetRandomEffectVariance() Gets the random effect variance-covariance matrix.
ModelMetadata<T> GetModelMetadata()
T ComputeICC() Computes the intraclass correlation coefficient (ICC).
T GetLogLikelihood(Matrix<T> x, Vector<T> y, int[] groupIndices) Gets the log-likelihood of the model.
Vector<T> GetFixedEffectStandardErrors() Gets the standard errors of fixed effects.
Vector<T> GetFixedEffects() Gets the fixed effect coefficients.
Vector<T> Predict(Matrix<T> input)
Vector<T> Predict(Matrix<T> input, int[] groupIndices) Predicts values for input samples with known group memberships.
void Train(Matrix<T> x, Vector<T> y)
void Train(Matrix<T> x, Vector<T> y, int[] groupIndices) Trains the mixed-effects model with explicit group indicators.