package learning

Ordering
  1. Alphabetic
Visibility
  1. Public
  2. All

Type Members

  1. abstract class EMTerminationCriteria extends AnyRef

    Termination criteria for EM algorithms.

    Termination criteria for EM algorithms. A termination criteria can be passed as an argument to the EM apply method.

  2. trait ExpectationMaximization extends Algorithm with ParameterLearner

    Expectation maximization iteratively produces an estimate of sufficient statistics for learnable parameters, then maximizes the parameters according to the estimate.

    Expectation maximization iteratively produces an estimate of sufficient statistics for learnable parameters, then maximizes the parameters according to the estimate. This trait can be extended with a different expectation or maximization algorithm; see the code for details.

  3. class ExpectationMaximizationWithFactors extends ExpectationMaximization

    An EM algorithm which learns parameters using a factored algorithm

  4. class GeneralizedEM extends ExpectationMaximization

    An EM algorithm which learns parameters using an inference algorithm provided as an argument

  5. class GeneralizedOnlineEM extends OnlineExpectationMaximization

    An EM algorithm which learns parameters using an inference algorithm provided as an argument

  6. class MaxIterations extends EMTerminationCriteria

    Terminate when the maximum number of iterations has been reached

  7. trait OnlineExpectationMaximization extends Online with ExpectationMaximization

    An EM algorithm which learns parameters incrementally

  8. class OnlineExpectationMaximizationWithFactors extends OnlineExpectationMaximization

    An online EM algorithm which learns parameters using a factored algorithm

  9. class SufficientStatisticsFactor extends AnyRef

    Methods for creating probabilistic factors associated with elements and their sufficient statistics.

  10. class SufficientStatisticsMagnitudes extends EMTerminationCriteria

    Terminate when the magnitude of sufficient statistics does not exhibit a change greater than the specified tolerance.

Ungrouped