Package weka.classifiers.evaluation
package weka.classifiers.evaluation
-
ClassDescriptionAbstract base class for pluggable classification/regression evaluation metrics.Subclass of Evaluation that provides a method for aggregating the results stored in another Evaluation object.Cells of this matrix correspond to counts of the number (or weight) of predictions for each actual value / predicted value combination.Generates points illustrating probablity cost tradeoffs that can be obtained by varying the threshold value between classes.Class for evaluating machine learning models.Helper routines for extracting metric values from built-in and plugin evaluation metrics.Contains utility functions for generating lists of predictions in various manners.An interface for information retrieval evaluation metrics to implement.Primarily a marker interface for information theoretic evaluation metrics to implement.Primarily a marker interface for interval-based evaluation metrics to implement.Generates points illustrating the prediction margin.Encapsulates an evaluatable nominal prediction: the predicted probability distribution plus the actual class value.Encapsulates an evaluatable numeric prediction: the predicted class value plus the actual class value.Encapsulates a single evaluatable prediction: the predicted value plus the actual class value.Analyzes linear regression model by using the Student's t-test on each coefficient.Primarily a marker interface for a "standard" evaluation metric - i.e.Generates points illustrating prediction tradeoffs that can be obtained by varying the threshold value between classes.Some evaluation measures may optimize thresholds on the class probabilities.Encapsulates performance functions for two-class problems.