Class LogisticRegression

java.lang.Object

org.episteme.core.mathematics.ml.LogisticRegression

public class LogisticRegression
extends Object

Logistic Regression - binary classification using sigmoid function.

Model: P(y=1|x) = 1/(1 + e^(-w·x)) Training: Gradient descent on log-likelihood Applications: Classification, probability estimation

Since:

1.0

Author:

Silvere Martin-Michiellot, Gemini AI (Google DeepMind)

Constructor Summary

Constructors

Constructor	Description
LogisticRegression()

Method Summary

Modifier and Type	Method	Description
void	fit(Real[][] X, Real[] y)	Trains logistic regression model.
void	fitWithRegularization(Real[][] X, Real[] y, Real lambda)	L2 regularization (Ridge): adds λ||w||² to loss.
Real	getBias()
Real[]	getWeights()
Real	predict(Real[] x)	Predicts class (0 or 1) using threshold 0.5.
Real	predictProba(Real[] x)	Predicts probability P(y=1|x).
Real	score(Real[][] X, Real[] y)	Computes accuracy on test set.
void	setLearningRate(Real lr)
void	setMaxIterations(int iter)

Methods inherited from class Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

LogisticRegression

public LogisticRegression()

Method Details

fit

public void fit(Real[][] X,
 Real[] y)

Trains logistic regression model.

Parameters:

X - training data (n samples × d features)

y - binary labels (0 or 1)

predictProba

public Real predictProba(Real[] x)

Predicts probability P(y=1|x).

predict

public Real predict(Real[] x)

Predicts class (0 or 1) using threshold 0.5.

score

public Real score(Real[][] X,
 Real[] y)

Computes accuracy on test set.

fitWithRegularization

public void fitWithRegularization(Real[][] X,
 Real[] y,
 Real lambda)

L2 regularization (Ridge): adds λ||w||² to loss.

setLearningRate

public void setLearningRate(Real lr)

setMaxIterations

public void setMaxIterations(int iter)

getWeights

public Real[] getWeights()

getBias

public Real getBias()