GridGain Developers Hub

Logistic Regression

Binary classification

Binary Logistic Regression is a special type of regression where a binary response variable is related to a set of explanatory variables, which can be discrete and/or continuous. The important point here to note is that in linear regression, the expected values of the response variable are modeled based on a combination of values taken by the predictors. In logistic regression Probability or Odds of the response taking a particular value is modeled based on the combination of values taken by the predictors. In the Apache Ignite ML module it is implemented via LogisticRegressionModel that solves the binary classification problem. It is a linear method with the loss function in the formulation given by the logistic loss:

mllog01

For binary classification problems, the algorithm outputs a binary logistic regression model. Given a new data point, denoted by x, the model makes predictions by applying the logistic function:

mllog02

By default, if f(wTx)>0.5 or \mathrm{f}(\wv^T x) > 0.5 (Tex formula), the outcome is positive, or negative otherwise. However, unlike linear SVMs, the raw output of the logistic regression model f(z) has a probabilistic interpretation (i.e., the probability that it is positive).

Multi-class classification

Multiclass Logistic Regression aims to assign labels to instances by using binary logistic regression, where the labels are drawn from a finite set of several elements. The implemented approach for doing so is to reduce the single multiclass problem into multiple binary classification problems via one-versus-all. The one-versus-all approach is the process of building binary classifiers which distinguish between one of the labels and the rest.

Model

The model keeps the pairs <ClassLabel, LogisticRegressionModel> and it enables a prediction to be made for a given vector of features, in the following way:

LogRegressionMultiClassModel
mdl = 

double prediction = mdl.withRawLabels(true).withThreshold(0.5).apply(observation)

Ignite supports these parameters for LogisticRegressionModel:

  • isKeepingRawLabels - controls the output label format: 0 and 1 for false value and raw distances from the separating hyperplane otherwise (default value: false)

  • threshold - a threshold to assign label 1 to the observation if the raw value is more than this threshold (default value: 0.5)

LogisticRegressionSGDTrainer
mdl = 

double prediction = mdl.withRawLabels(true).withThreshold(0.5).apply(observation)

Trainer

Trainer of the multi-class logistic regression model runs a number of binary logistic regression trainers under the hood.

Ignite supports the following parameters for LogRegressionMultiClassTrainer:

  • updatesStgy - update strategy.

  • maxIterations - max amount of iterations before convergence.

  • batchSize - The size of learning batch.

  • locIterations - the amount of local iterations of SGD algorithm.

  • seed - seed value for internal random purposes to reproduce training results.

LogRegressionMultiClassTrainer<?> trainer = new LogRegressionMultiClassTrainer<>()
  .withUpdatesStgy(UPDATES_STRATEGY)
  .withAmountOfIterations(MAX_ITERATIONS)
  .withAmountOfLocIterations(BATCH_SIZE)
  .withBatchSize(LOC_ITERATIONS)
  .withSeed(SEED);

// Build the model
LogisticRegressionModel mdl = trainer.fit(
  ignite,
  dataCache,
  featureExtractor,
  labelExtractor
);

All properties will be propagated for each pair one-versus-all LogRegressionMultiClassTrainer.