Package smile.classification

Class LogisticRegression

java.lang.Object
smile.classification.AbstractClassifier<double[]>
smile.classification.LogisticRegression
All Implemented Interfaces:
Serializable, ToDoubleFunction<double[]>, ToIntFunction<double[]>, Classifier<double[]>
Direct Known Subclasses:
LogisticRegression.Binomial, LogisticRegression.Multinomial
public abstract class LogisticRegression extends AbstractClassifier<double[]>
Logistic regression. Logistic regression (logit model) is a generalized linear model used for binomial regression. Logistic regression applies maximum likelihood estimation after transforming the dependent into a logit variable. A logit is the natural log of the odds of the dependent equaling a certain value or not (usually 1 in binary logistic models, the highest value in multinomial models). In this way, logistic regression estimates the odds of a certain event (value) occurring.

Goodness-of-fit tests such as the likelihood ratio test are available as indicators of model appropriateness, as is the Wald statistic to test the significance of individual independent variables.

Logistic regression has many analogies to ordinary least squares (OLS) regression. Unlike OLS regression, however, logistic regression does not assume linearity of relationship between the raw values of the independent variables and the dependent, does not require normally distributed variables, does not assume homoscedasticity, and in general has less stringent requirements.

Compared with linear discriminant analysis, logistic regression has several advantages:

  • It is more robust: the independent variables don't have to be normally distributed, or have equal variance in each group
  • It does not assume a linear relationship between the independent variables and dependent variable.
  • It may handle nonlinear effects since one can add explicit interaction and power terms.
However, it requires much more data to achieve stable, meaningful results.

Logistic regression also has strong connections with neural network and maximum entropy modeling. For example, binary logistic regression is equivalent to a one-layer, single-output neural network with a logistic activation function trained under log loss. Similarly, multinomial logistic regression is equivalent to a one-layer, softmax-output neural network.

Logistic regression estimation also obeys the maximum entropy principle, and thus logistic regression is sometimes called "maximum entropy modeling", and the resulting classifier the "maximum entropy classifier".

See Also:

  • Constructor Details

    • LogisticRegression

      public LogisticRegression(int p, double L, double lambda, IntSet labels)
      Constructor.
      Parameters:
      p - the dimension of input data.
      L - the log-likelihood of learned model.
      lambda - lambda > 0 gives a "regularized" estimate of linear weights which often has superior generalization performance, especially when the dimensionality is high.
      labels - the class label encoder.

  • Method Details

    • binomial

      public static LogisticRegression.Binomial binomial(double[][] x, int[] y)
      Fits binomial logistic regression.
      Parameters:
      x - training samples.
      y - training labels.
      Returns:
      the model.

    • binomial

      public static LogisticRegression.Binomial binomial(double[][] x, int[] y, Properties params)
      Fits binomial logistic regression.
      Parameters:
      x - training samples.
      y - training labels.
      params - the hyperparameters.
      Returns:
      the model.

    • binomial

      public static LogisticRegression.Binomial binomial(double[][] x, int[] y, double lambda, double tol, int maxIter)
      Fits binomial logistic regression.
      Parameters:
      x - training samples.
      y - training labels.
      lambda - lambda > 0 gives a "regularized" estimate of linear weights which often has superior generalization performance, especially when the dimensionality is high.
      tol - the tolerance for stopping iterations.
      maxIter - the maximum number of iterations.
      Returns:
      the model.

    • multinomial

      public static LogisticRegression.Multinomial multinomial(double[][] x, int[] y)
      Fits multinomial logistic regression.
      Parameters:
      x - training samples.
      y - training labels.
      Returns:
      the model.

    • multinomial

      public static LogisticRegression.Multinomial multinomial(double[][] x, int[] y, Properties params)
      Fits multinomial logistic regression.
      Parameters:
      x - training samples.
      y - training labels.
      params - the hyperparameters.
      Returns:
      the model.

    • multinomial

      public static LogisticRegression.Multinomial multinomial(double[][] x, int[] y, double lambda, double tol, int maxIter)
      Fits multinomial logistic regression.
      Parameters:
      x - training samples.
      y - training labels.
      lambda - lambda > 0 gives a "regularized" estimate of linear weights which often has superior generalization performance, especially when the dimensionality is high.
      tol - the tolerance for stopping iterations.
      maxIter - the maximum number of iterations.
      Returns:
      the model.

    • fit

      public static LogisticRegression fit(double[][] x, int[] y)
      Fits logistic regression.
      Parameters:
      x - training samples.
      y - training labels.
      Returns:
      the model.

    • fit

      public static LogisticRegression fit(double[][] x, int[] y, Properties params)
      Fits logistic regression.
      Parameters:
      x - training samples.
      y - training labels.
      params - the hyperparameters.
      Returns:
      the model.

    • fit

      public static LogisticRegression fit(double[][] x, int[] y, double lambda, double tol, int maxIter)
      Fits logistic regression.
      Parameters:
      x - training samples.
      y - training labels.
      lambda - lambda > 0 gives a "regularized" estimate of linear weights which often has superior generalization performance, especially when the dimensionality is high.
      tol - the tolerance to stop iterations.
      maxIter - the maximum number of iterations.
      Returns:
      the model.

    • soft

      public boolean soft()
      Description copied from interface: Classifier
      Returns true if this is a soft classifier that can estimate the posteriori probabilities of classification.
      Returns:
      true if soft classifier.

    • online

      public boolean online()
      Description copied from interface: Classifier
      Returns true if this is an online learner.
      Returns:
      true if online learner.

    • setLearningRate

      public void setLearningRate(double rate)
      Sets the learning rate of stochastic gradient descent. It is a good practice to adapt the learning rate for different data sizes. For example, it is typical to set the learning rate to eta/n, where eta is in [0.1, 0.3] and n is the size of the training data.
      Parameters:
      rate - the learning rate.

    • getLearningRate

      public double getLearningRate()
      Returns the learning rate of stochastic gradient descent.
      Returns:
      the learning rate of stochastic gradient descent.

    • loglikelihood

      public double loglikelihood()
      Returns the log-likelihood of model.
      Returns:
      the log-likelihood of model.

    • AIC

      public double AIC()
      Returns the AIC score.
      Returns:
      the AIC score.