Package

axle

ml

Permalink

package ml

Visibility
  1. Public
  2. All

Type Members

  1. case class ClassifierPerformance[N, DATA, F, G](data: F, retrieve: (DATA) ⇒ Boolean, relevant: (DATA) ⇒ Boolean)(implicit functor: Functor[F, DATA, (N, N, N, N), G], agg: Aggregatable[G, (N, N, N, N), (N, N, N, N)], field: Field[N]) extends Product with Serializable

    Permalink

    ClassifierPerformance computes measures of classification performance

    ClassifierPerformance computes measures of classification performance

    They are:

    * Precision * Recall * Specificity * Accuracy * F1

    The (boolean) "classification task" is defined by two arguments:

    1) predict: given a datum, determines whether the value is "in" the retrieved set 2) actual : given a datum, determines whether the value is *actually* "in" the retrieved set

    See http://en.wikipedia.org/wiki/Precision_and_recall for more information.

    http://en.wikipedia.org/wiki/F1_score

  2. case class ConfusionMatrix[T, CLASS, L, F, M, G, H](classifier: (T) ⇒ CLASS, data: F, labelExtractor: (T) ⇒ L, classes: IndexedSeq[CLASS])(implicit evidence$1: Order[CLASS], evidence$2: Order[L], la: LinearAlgebra[M, Int, Int, Double], finite: Finite[F, Int], functorF: Functor[F, T, (L, CLASS), G], functorG: Functor[G, (L, CLASS), L, H], sf: SetFrom[H, L], mr: MapReducible[G, (L, CLASS), Int, (Int, CLASS), Map[(Int, CLASS), Int]], mf: MapFrom[List[(L, Int)], L, Int]) extends Product with Serializable

    Permalink
  3. case class GeneticAlgorithm[G <: HList, Z <: HList](populationSize: Int = 1000, numGenerations: Int = 100)(implicit species: Species[G], zipper: Aux[::[G, ::[G, HNil]], Z], mapperMix: Mapper[Mixer.type, Z], mapperMutate: Mapper[Mutator.type, Z]) extends Product with Serializable

    Permalink
  4. case class GeneticAlgorithmLog[G](winners: IndexedSeq[G], mins: TreeMap[Int, Double], maxs: TreeMap[Int, Double], aves: TreeMap[Int, Double]) extends Product with Serializable

    Permalink
  5. class HMMEdge[N] extends AnyRef

    Permalink
  6. case class HiddenMarkovModel[DG](states: IndexedSeq[UnobservableMarkovModelState], observations: IndexedSeq[ObservableMarkovModelState], startProbability: Map[UnobservableMarkovModelState, Double], transitionProbability: Map[UnobservableMarkovModelState, Map[UnobservableMarkovModelState, Double]], emissionProbability: Map[UnobservableMarkovModelState, Map[ObservableMarkovModelState, Double]])(implicit dg: DirectedGraph[DG, MarkovModelState, HMMEdge[Double]]) extends Product with Serializable

    Permalink
  7. case class IdentityFeatureNormalizer[M](X: M)(implicit la: LinearAlgebra[M, Int, Int, Double]) extends Normalize[M] with Product with Serializable

    Permalink
  8. case class KMeans[T, F, G, M](data: F, N: Int, featureExtractor: (T) ⇒ Seq[Double], normalizerMaker: (M) ⇒ Normalize[M], K: Int, iterations: Int)(implicit evidence$1: Eq[T], space: MetricSpace[M, Double], functor: Functor[F, T, Seq[Double], G], la: LinearAlgebra[M, Int, Int, Double], index: Indexed[G, Int, Seq[Double]], finite: Finite[F, Int]) extends (T) ⇒ Int with Product with Serializable

    Permalink

    KMeans

    KMeans

    T

    type of the objects being classified

  9. case class LatentSemanticAnalysis() extends Product with Serializable

    Permalink
  10. case class LinearFeatureNormalizer[M](X: M)(implicit la: LinearAlgebra[M, Int, Int, Double]) extends Normalize[M] with Product with Serializable

    Permalink
  11. case class LinearRegression[D, M](examples: Seq[D], numFeatures: Int, featureExtractor: (D) ⇒ Seq[Double], objectiveExtractor: (D) ⇒ Double, α: Double = 0.1, iterations: Int = 100)(implicit la: LinearAlgebra[M, Int, Int, Double]) extends (D) ⇒ Double with Product with Serializable

    Permalink
  12. case class LogisticRegression[D, M](examples: List[D], numFeatures: Int, featureExtractor: (D) ⇒ List[Double], objectiveExtractor: (D) ⇒ Boolean, α: Double = 0.1, numIterations: Int = 100)(implicit la: LinearAlgebra[M, Int, Int, Double]) extends (List[Double]) ⇒ Double with Product with Serializable

    Permalink
  13. case class MarkovModelStartState() extends MarkovModelState with Product with Serializable

    Permalink
  14. sealed trait MarkovModelState extends AnyRef

    Permalink

    http://en.wikipedia.org/wiki/Hidden_Markov_model

  15. case class NaiveBayesClassifier[DATA, FEATURE, CLASS, F, G, N](data: F, featureRandomVariables: List[Distribution[FEATURE, N]], classRandomVariable: Distribution[CLASS, N], featureExtractor: (DATA) ⇒ List[FEATURE], classExtractor: (DATA) ⇒ CLASS)(implicit evidence$1: Order[FEATURE], evidence$2: Order[CLASS], evidence$3: Eq[CLASS], evidence$4: Field[N], evidence$5: Order[N], agg: Aggregatable[F, DATA, Map[(CLASS, String, FEATURE), N]], functor: Functor[F, DATA, CLASS, G], tal: Talliable[G, CLASS, N]) extends (DATA) ⇒ CLASS with Product with Serializable

    Permalink
  16. trait Normalize[M] extends (Seq[Double]) ⇒ M

    Permalink
  17. case class ObservableMarkovModelState(label: String) extends MarkovModelState with Product with Serializable

    Permalink
  18. case class PCAFeatureNormalizer[M](cutoff: Double, X: M)(implicit la: LinearAlgebra[M, Int, Int, Double]) extends Normalize[M] with Product with Serializable

    Permalink
  19. case class PrincipalComponentAnalysis() extends Product with Serializable

    Permalink
  20. trait Species[G] extends AnyRef

    Permalink
  21. case class UnobservableMarkovModelState(label: String) extends MarkovModelState with Product with Serializable

    Permalink
  22. case class ZScoreFeatureNormalizer[M](X: M)(implicit la: LinearAlgebra[M, Int, Int, Double]) extends Normalize[M] with Product with Serializable

    Permalink

Value Members

  1. object BaumWelch

    Permalink

    http://en.wikipedia.org/wiki/Baum%E2%80%93Welch_algorithm

    http://en.wikipedia.org/wiki/Baum%E2%80%93Welch_algorithm

    The Baum–Welch algorithm is a particular case of a generalized expectation-maximization (GEM) algorithm. It can compute maximum likelihood estimates and posterior mode estimates for the parameters (transition and emission probabilities) of an HMM, when given only emissions as training data.

  2. object ClassifierPerformance extends Serializable

    Permalink
  3. object ConfusionMatrix extends Serializable

    Permalink
  4. object KMeans extends Serializable

    Permalink
  5. object MarkovModelStartState extends Serializable

    Permalink
  6. object MarkovModelState

    Permalink
  7. object Mater extends Poly1

    Permalink
  8. object Mixer extends Poly1

    Permalink
  9. object Mutator extends Poly1

    Permalink
  10. object NaiveBayesClassifier extends Serializable

    Permalink
  11. object ObservableMarkovModelState extends Serializable

    Permalink
  12. object RankedClassifierPerformance

    Permalink
  13. object UnobservableMarkovModelState extends Serializable

    Permalink
  14. package distance

    Permalink

Ungrouped