LogNormalAFT<T>

Models & Types · AiDotNet.SurvivalAnalysis

Implements the Log-Normal Accelerated Failure Time (AFT) model for survival analysis.

Inheritance: SurvivalModelBase<T> → LogNormalAFT<T>

For Beginners

The Log-Normal AFT model assumes that the log of survival times follows a normal distribution. This is appropriate when survival times have a bell-curve shape after log-transformation, common in many biomedical and engineering applications.

How It Works

The Model: log(T) = β₀ + β₁X₁ + … + βₚXₚ + σε where ε ~ N(0,1) is standard normal.

Interpretation:

  • exp(β) gives the multiplicative effect on median survival time
  • A positive coefficient increases survival time (protective)
  • A negative coefficient decreases survival time (harmful)

When to use Log-Normal vs Weibull:

  • Log-Normal: Hazard first increases then decreases (non-monotonic)
  • Weibull: Hazard is monotonic (increasing, decreasing, or constant)

Reference: Lawless (2003), Statistical Models and Methods for Lifetime Data

Example

// age and treatment arm per patient
var features = new Matrix<double>(new double[,]
{
    { 45, 1 }, { 52, 0 }, { 38, 1 }, { 61, 0 }, { 47, 1 }, { 55, 0 }
});
// months each patient was observed
var times = new Vector<double>(new double[] { 5.0, 12.0, 3.0, 18.0, 9.0, 21.0 });
// 1 = the event happened; 0 = censored, still fine when the study ended or lost to follow-up
var events = new Vector<int>(new int[] { 1, 0, 1, 0, 1, 1 });
            
var result = new AiModelBuilder<double, Matrix<double>, Vector<double>>()
    .ConfigureModel(new LogNormalAFT<double>(maxIterations: 100, tolerance: 1e-6))
    .Build(features, times, events);
            
var risk = result.Predict(features);

Constructors

Constructor Summary
LogNormalAFT(int maxIterations, double tolerance) Creates a new Log-Normal AFT model.

Properties

Property Summary
T Intercept { get; set; } Gets the intercept term (β₀ = μ).
T Scale { get; set; } Gets the scale parameter (σ).
Vector<T> Coefficients { get; set; } Gets the regression coefficients (β).
double Tolerance { get; } Gets the tolerance for convergence.
int MaxIterations { get; } Gets the maximum iterations for optimization.

Methods

Method Summary
IFullModel<T, Matrix<T>, Vector<T>> WithParameters(Vector<T> parameters)
Matrix<T> PredictSurvivalProbability(Matrix<T> x, Vector<T> times) Predicts survival probabilities at specified times.
Vector<T> GetBaselineSurvival(Vector<T> times) Gets the baseline survival function.
Vector<T> Predict(Matrix<T> input) Predicts median survival time.
Vector<T> PredictHazardRatio(Matrix<T> x) Predicts acceleration factors (hazard ratios at baseline hazard).