Interface StatusAnalyticsModel


public interface StatusAnalyticsModel
  • Method Details

    • learn

      void learn(Stream<Double[]> features, Stream<Double> labels)
      Train model with provided observations (features, labels/targets)
      Parameters:
      features - Stream of feature observation values
      labels - target observation values
    • predict

      Double predict(Double[] feature)
      Return a prediction given observation values
      Parameters:
      feature - feature observation values values
      Returns:
      prediction of target/label
    • predictVariable

      Double predictVariable(Integer predictVariableIndex, Map<Integer,Double> knownVariablesWithIndex, Double label)
      Predict a feature given a known target and known predictor values (if multiple predictors are included with model)
      Parameters:
      predictVariableIndex - index of feature that we would like to predict (index should align with order provided in model learn method)
      knownVariablesWithIndex - a map of known predictor values with their indexes if available
      label - known target value
      Returns:
      prediction for variable
    • supportsOnlineLearning

      Boolean supportsOnlineLearning()
      Indicate if model supports online learning (e.g. can learn new observation samples to create a model)
      Returns:
      boolean indicating online learning support
    • getScores

      Map<String,Double> getScores()
      Returns a map of scores relevant to model (e.g. rSquared, Confidence Intervals, etc.)
      Returns:
      Map of score names with values
    • clear

      void clear()
      Resets a model by clearing observations and other calculations