MultinomialLogisticRegression<T>

Models & Types · AiDotNet.Regression

Represents a multinomial logistic regression model for multi-class classification problems.

Inheritance: RegressionBase<T> → MultinomialLogisticRegression<T>

Implements: IGeneratedParameterRegistrar<T>

For Beginners

Multinomial logistic regression is a method for classifying data into multiple categories.

Think of it like a voting system where:

  • Each feature (input variable) gets to “vote” for different categories
  • The importance of each feature’s vote is learned from training data
  • For any new data point, we count the weighted votes for each category
  • The category with the most votes wins and becomes the prediction

For example, when classifying emails into categories like “work,” “personal,” or “spam,” certain words might strongly suggest one category over others. The model learns which features (words) are most helpful for distinguishing between the different categories.

How It Works

Multinomial logistic regression extends binary logistic regression to handle multiple classes. It models the probabilities of different possible outcomes using the softmax function. For each class, the model learns a set of coefficients that determine how each feature affects the probability of that class. During prediction, it assigns the input to the class with the highest probability.

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 MultinomialLogisticRegression<double>())
    .ConfigureDataLoader(DataLoaders.FromArrays(features, targets))
    .BuildAsync();

Console.WriteLine("Trained MultinomialLogisticRegression.");

Constructors

Constructor Summary
MultinomialLogisticRegression(MultinomialLogisticRegressionOptions<T> options, IRegularization<T, Matrix<T>, Vector<T>> regularization) Initializes a new instance of the MultinomialLogisticRegression class with optional custom options and regularization.

Properties

Property Summary
IReadOnlyList<T> ClassLabels { get; } The class labels the model was trained on, ascending.
bool SupportsParameterInitialization { get; } Multinomial logistic is a classification model — no optimizer parameter injection.

Methods

Method Summary
Matrix<T> PredictProbabilities(Matrix<T> x) Predicts the probabilities of each class for new data points.
ModelOptions GetOptions()
Vector<T> Predict(Matrix<T> x) Predicts the class labels for new data points using the trained multinomial logistic regression model.
void Train(Matrix<T> x, Vector<T> y) Trains the multinomial logistic regression model using the provided features and target values.