Packages

class TwoBAN_OS0 extends BayesClassifier

The TwoBAN_OS0 class implements an Integer-Based Bayesian Network Classifier, which is a commonly used such classifier for discrete input data. Each node is limited to have at most 2 parents, and hence the "2" in the class name TwoBAN_OS. The classifier is trained using a data matrix 'x' and a classification vector 'y'. Each data vector in the matrix is classified into one of 'k' classes numbered 0, ..., k-1. Prior probabilities are calculated based on the population of each class in the training-set. Relative posterior probabilities are computed by multiplying these by values computed using conditional probabilities. The classifier supports limited dependency between features/variables.

This classifier uses the standard cross-validation technique. ------------------------------------------------------------------------------

Linear Supertypes
Known Subclasses
Ordering
  1. Alphabetic
  2. By Inheritance
Inherited
  1. TwoBAN_OS0
  2. BayesClassifier
  3. BayesMetrics
  4. ClassifierInt
  5. Error
  6. Classifier
  7. AnyRef
  8. Any
  1. Hide All
  2. Show All
Visibility
  1. Public
  2. All

Instance Constructors

  1. new TwoBAN_OS0(x: MatriI, y: VectoI, fn: Array[String], k: Int, cn: Array[String], vc: Array[Int] = null, thres: Double = 0.0, me: Double = me_default, PARALLELISM: Int = ...)

    x

    the integer-valued data vectors stored as rows of a matrix

    y

    the class vector, where y(l) = class for row l of the matrix, x(l)

    fn

    the names for all features/variables

    k

    the number of classes

    cn

    the names for all classes

    vc

    the value count (number of distinct values) for each feature

    thres

    the correlation threshold between 2 features for possible parent-child relationship

    me

    use m-estimates (me == 0 => regular MLE estimates)

    PARALLELISM

    the level of parallelism, the number of threads to use

Value Members

  1. final def !=(arg0: Any): Boolean
    Definition Classes
    AnyRef → Any
  2. final def ##(): Int
    Definition Classes
    AnyRef → Any
  3. final def ==(arg0: Any): Boolean
    Definition Classes
    AnyRef → Any
  4. val N0: Double
    Attributes
    protected
    Definition Classes
    BayesClassifier
  5. var additive: Boolean
    Attributes
    protected
    Definition Classes
    BayesClassifier
  6. def aic(vc: Array[Int], vcp1: VectoI, vcp2: VectoI, popX: HMatrix5[Int], k: Int, me: Float = me_default): Double

    Compute the 'AIC' for the given Bayesian Network structure and data.

    Compute the 'AIC' for the given Bayesian Network structure and data.

    vc

    the value count

    vcp1

    the value count for parent 1

    vcp2

    the value count for parent 2

    popX

    the population counts

    k

    the number of classes

    me

    the m-estimate value

    Definition Classes
    BayesMetrics
  7. final def asInstanceOf[T0]: T0
    Definition Classes
    Any
  8. def calcCMI(idx: IndexedSeq[Int], vca: Array[Int]): MatrixD

    Compute the conditional mutual information matrix

    Compute the conditional mutual information matrix

    Definition Classes
    BayesClassifier
  9. def calcCorrelation: MatriD

    Calculate the correlation matrix for the feature vectors 'fea'.

    Calculate the correlation matrix for the feature vectors 'fea'. If the correlations are too high, the independence assumption may be dubious.

    Definition Classes
    ClassifierInt
  10. def calcCorrelation2(zrg: Range, xrg: Range): MatriD

    Calculate the correlation matrix for the feature vectors of Z (Level 3) and those of X (level 2).

    Calculate the correlation matrix for the feature vectors of Z (Level 3) and those of X (level 2). If the correlations are too high, the independence assumption may be dubious.

    zrg

    the range of Z-columns

    xrg

    the range of X-columns

    Definition Classes
    ClassifierInt
  11. def classify(z: VectoI): (Int, String, Double)

    Given a discrete data vector 'z', classify it returning the class number (0, ..., k-1) with the highest relative posterior probability.

    Given a discrete data vector 'z', classify it returning the class number (0, ..., k-1) with the highest relative posterior probability. Return the best class, its name and its realtive probability

    z

    the data vector to classify

    Definition Classes
    TwoBAN_OS0Classifier
  12. def classify(xx: MatriI): VectoI

    Classify all of the row vectors in matrix 'xx'.

    Classify all of the row vectors in matrix 'xx'.

    xx

    the row vectors to classify

    Definition Classes
    ClassifierInt
  13. def classify(z: VectoD): (Int, String, Double)

    Given a new continuous data vector 'z', determine which class it fits into, returning the best class, its name and its relative probability.

    Given a new continuous data vector 'z', determine which class it fits into, returning the best class, its name and its relative probability. Override in classes that require precise real values for classification.

    z

    the real vector to classify

    Definition Classes
    ClassifierIntClassifier
  14. def clone(): AnyRef
    Attributes
    protected[java.lang]
    Definition Classes
    AnyRef
    Annotations
    @native() @throws( ... )
  15. def cmiJoint(p_C: VectorD, p_CX: HMatrix3[Double], p_CXZ: HMatrix5[Double]): MatrixD

    Compute conditional mutual information matrix given the marginal probability of C and joint probabilities of CXZ and CX, where C is the class (parent), and X & Z are features.

    Compute conditional mutual information matrix given the marginal probability of C and joint probabilities of CXZ and CX, where C is the class (parent), and X & Z are features.

    p_C

    the marginal probability of C

    p_CX

    the joint probability of C and X

    p_CXZ

    the joint probability of C, X, and Z

    Definition Classes
    BayesClassifier
    See also

    en.wikipedia.org/wiki/Conditional_mutual_information

  16. def computeParent(featureOrder: VectoI = featureOrder, cmiMx: MatrixD): MatrixI

    Compute the parent of each feature based on the correlation matrix.

    Compute the parent of each feature based on the correlation matrix. Feature 'x_i' is only a possible candidate for parent of feature 'x_j' if 'x_i' appears before 'x_j' in 'featureOrder'.

    featureOrder

    order of the features

    cmiMx

    the Conditional Mutual Information matrix

  17. def computeVcp(): Unit

    Compute the value counts of each parent feature based on the parent matrix.

    Compute the value counts of each parent feature based on the parent matrix. Let 1 be the default value count when there is no parent.

  18. def crossValidate(nx: Int = 10, show: Boolean = false): Double

    Test the accuracy of the classified results by cross-validation, returning the accuracy.

    Test the accuracy of the classified results by cross-validation, returning the accuracy. The "test data" starts at 'testStart' and ends at 'testEnd', the rest of the data is "training data'. FIX - should return a StatVector

    nx

    the number of crosses and cross-validations (defaults to 10x).

    show

    the show flag (show result from each iteration)

    Definition Classes
    Classifier
  19. def crossValidateRand(nx: Int = 10, show: Boolean = false): Double

    Test the accuracy of the classified results by cross-validation, returning the accuracy.

    Test the accuracy of the classified results by cross-validation, returning the accuracy. This version of cross-validation relies on "subtracting" frequencies from the previously stored global data to achieve efficiency. FIX - are the comments correct? FIX - should return a StatVector

    nx

    number of crosses and cross-validations (defaults to 10x).

    show

    the show flag (show result from each iteration)

    Definition Classes
    Classifier
  20. final def eq(arg0: AnyRef): Boolean
    Definition Classes
    AnyRef
  21. def equals(arg0: Any): Boolean
    Definition Classes
    AnyRef → Any
  22. val f_C: VectorI
    Attributes
    protected
    Definition Classes
    BayesClassifier
  23. var f_CX: HMatrix3[Int]
    Attributes
    protected
    Definition Classes
    BayesClassifier
  24. var f_CXPP: HMatrix5[Int]
    Attributes
    protected
  25. var f_CXZ: HMatrix5[Int]
    Attributes
    protected
    Definition Classes
    BayesClassifier
  26. var f_X: HMatrix2[Int]
    Attributes
    protected
    Definition Classes
    BayesClassifier
  27. val featOrderArray: Array[VectoI]
    Attributes
    protected
  28. var featureOrder: VectoI
    Attributes
    protected
  29. def featureSelection(TOL: Double = 0.01): Unit

    Perform feature selection on the classifier.

    Perform feature selection on the classifier. Use backward elimination technique, that is, remove the least significant feature, in terms of cross- validation accuracy, in each round.

    TOL

    tolerance indicating negligible accuracy loss when removing features

    Definition Classes
    ClassifierInt
  30. def finalize(): Unit
    Attributes
    protected[java.lang]
    Definition Classes
    AnyRef
    Annotations
    @throws( classOf[java.lang.Throwable] )
  31. def fit(y: VectoI, yp: VectoI, k: Int = 2): VectoD

    Return the quality of fit including 'acc', 'prec', 'recall', 'kappa'.

    Return the quality of fit including 'acc', 'prec', 'recall', 'kappa'. Override to add more quality of fit measures.

    y

    the actual class labels

    yp

    the precicted class labels

    k

    the number of class labels

    Definition Classes
    Classifier
    See also

    ConfusionMat

    medium.com/greyatom/performance-metrics-for-classification-problems-in-machine-learning-part-i-b085d432082b

  32. def fitLabel: Seq[String]

    Return the labels for the fit.

    Return the labels for the fit. Override when necessary.

    Definition Classes
    Classifier
  33. final def flaw(method: String, message: String): Unit
    Definition Classes
    Error
  34. val fset: Array[Boolean]

    the set of features to turn on or off.

    the set of features to turn on or off. All features are on by default. Used for feature selection.

    Attributes
    protected
    Definition Classes
    ClassifierInt
  35. final def getClass(): Class[_]
    Definition Classes
    AnyRef → Any
    Annotations
    @native()
  36. def getFeatureOrder: VectoI

    Return the feature order.

  37. def getParent: MatrixI

    Return the parent.

  38. def hashCode(): Int
    Definition Classes
    AnyRef → Any
    Annotations
    @native()
  39. final def isInstanceOf[T0]: Boolean
    Definition Classes
    Any
  40. def learnStructure(cmiMx: MatrixD): Unit

    Learn the structure of Bayes Networks2 classier model by using the 'CMI' criterion.

    Learn the structure of Bayes Networks2 classier model by using the 'CMI' criterion. Limited dependencies between variables/features are also supported. Maximum number of parents for each feature is 2.

  41. def logLikelihood(vc: Array[Int], vcp1: VectoI, vcp2: VectoI, popX: HMatrix5[Int], k: Int, me: Float = me_default): Double

    Compute the Log-Likelihood for the given Bayesian Network structure and data.

    Compute the Log-Likelihood for the given Bayesian Network structure and data.

    vc

    the value count

    vcp1

    the value count for parent 1

    vcp2

    the value count for parent 2

    popX

    the population counts

    k

    the number of classes

    me

    the m-estimate value

    Definition Classes
    BayesMetrics
  42. val m: Int

    the number of data vectors in training/test-set (# rows)

    the number of data vectors in training/test-set (# rows)

    Attributes
    protected
    Definition Classes
    ClassifierInt
  43. val maxRandomRestarts: Int
    Attributes
    protected
  44. val md: Double

    the training-set size as a Double

    the training-set size as a Double

    Attributes
    protected
    Definition Classes
    ClassifierInt
  45. val n: Int

    the number of features/variables (# columns)

    the number of features/variables (# columns)

    Attributes
    protected
    Definition Classes
    ClassifierInt
  46. val nd: Double

    the feature-set size as a Double

    the feature-set size as a Double

    Attributes
    protected
    Definition Classes
    ClassifierInt
  47. final def ne(arg0: AnyRef): Boolean
    Definition Classes
    AnyRef
  48. final def notify(): Unit
    Definition Classes
    AnyRef
    Annotations
    @native()
  49. final def notifyAll(): Unit
    Definition Classes
    AnyRef
    Annotations
    @native()
  50. var p_C: VectorD
    Attributes
    protected
    Definition Classes
    BayesClassifier
  51. val p_X_CPP: HMatrix5[Double]
    Attributes
    protected
  52. var parent: MatrixI
    Attributes
    protected
  53. val permutedVec: PermutedVecI
    Attributes
    protected
  54. def reset(): Unit

    Reset or re-initialize the frequency tables to 0.

    Reset or re-initialize the frequency tables to 0.

    Definition Classes
    TwoBAN_OS0Classifier
  55. val scoreArray: VectorD
    Attributes
    protected
  56. def scoreCMI(parent: MatrixI = parent, cmiMx: MatrixD): Double

    Sum up the conditional mutual information score of the BN structure represented by 'parent'

    Sum up the conditional mutual information score of the BN structure represented by 'parent'

    parent

    parent/predecessor matrix (dim = n x 2)

    cmiMx

    the conditional mutual information matrix

  57. def shiftToZero(): Unit

    Shift the 'x' Matrix so that the minimum value for each column equals zero.

    Shift the 'x' Matrix so that the minimum value for each column equals zero.

    Definition Classes
    ClassifierInt
  58. def size: Int

    Return the number of data vectors/points in the entire dataset (training + testing),

    Return the number of data vectors/points in the entire dataset (training + testing),

    Definition Classes
    ClassifierIntClassifier
  59. var smooth: Boolean
    Attributes
    protected
    Definition Classes
    BayesClassifier
  60. def split(indices: IndexedSeq[Int], k: Int): Array[IndexedSeq[Int]]

    Split 'indices' into 'k' arrays of equal sizes (perhaps except for the last one)

    Split 'indices' into 'k' arrays of equal sizes (perhaps except for the last one)

    indices

    the ParSeq to be splitted

    k

    the number of pieces the vector is to be splitted

    Definition Classes
    BayesClassifier
  61. final def synchronized[T0](arg0: ⇒ T0): T0
    Definition Classes
    AnyRef
  62. def test(itest: Array[Int]): Double

    Test the quality of the training with a test-set and return the fraction of correct classifications.

    Test the quality of the training with a test-set and return the fraction of correct classifications.

    itest

    indices of the instances considered test data

    Definition Classes
    BayesClassifier
  63. def test(xx: MatriI, yy: VectoI): Double

    Test the quality of the training with a test-set and return the fraction of correct classifications.

    Test the quality of the training with a test-set and return the fraction of correct classifications.

    xx

    the integer-valued test vectors stored as rows of a matrix

    yy

    the test classification vector, where 'yy_i = class' for row 'i' of 'xx'

    Definition Classes
    ClassifierInt
  64. def test(itest: IndexedSeq[Int]): Double

    Test the quality of the training with a test-set and return the fraction of correct classifications.

    Test the quality of the training with a test-set and return the fraction of correct classifications.

    itest

    indices of the instances considered test data

    Definition Classes
    ClassifierIntClassifier
  65. def test(testStart: Int, testEnd: Int): Double

    Test the quality of the training with a test dataset and return the fraction of correct classifications.

    Test the quality of the training with a test dataset and return the fraction of correct classifications. Can be used when the dataset is randomized so that the testing/training part of a dataset corresponds to simple slices of vectors and matrices.

    testStart

    the beginning of test region (inclusive).

    testEnd

    the end of test region (exclusive).

    Definition Classes
    Classifier
  66. val tiny: Double
    Attributes
    protected
    Definition Classes
    BayesClassifier
  67. def toString(): String
    Definition Classes
    AnyRef → Any
  68. def toggleSmooth(): Unit

    Toggle the value of the 'smooth' property.

    Toggle the value of the 'smooth' property.

    Definition Classes
    BayesClassifier
  69. def train(itest: IndexedSeq[Int]): TwoBAN_OS0

    Train the classifier by computing the probabilities for C, and the conditional probabilities for X_j.

    Train the classifier by computing the probabilities for C, and the conditional probabilities for X_j. This is the quick version that uses the "subtraction" method to achieve efficiency.

    itest

    indices of the instances considered testing data

    Definition Classes
    TwoBAN_OS0Classifier
  70. def train(): Classifier

    Train the classifier by computing the probabilities from a training dataset of data vectors and their classifications.

    Train the classifier by computing the probabilities from a training dataset of data vectors and their classifications. Must be implemented in any extending class. Can be used when the whole dataset is used for training.

    Definition Classes
    Classifier
  71. def train(testStart: Int, testEnd: Int): Classifier

    Train the classifier by computing the probabilities from a training dataset of data vectors and their classifications.

    Train the classifier by computing the probabilities from a training dataset of data vectors and their classifications. Must be implemented in any extending class. Can be used when the dataset is randomized so that the training part of a dataset corresponds to simple slices of vectors and matrices.

    testStart

    starting index of test region (inclusive) used in cross-validation

    testEnd

    ending index of test region (exclusive) used in cross-validation

    Definition Classes
    Classifier
  72. def updateFreq(i: Int, f_C: VectoI, f_X: HMatrix2[Int], f_CX: HMatrix3[Int], f_CXZ: HMatrix5[Int]): Unit

    Increment frequency counters used in CMI calculations based on the 'i'th row of the data matrix.

    Increment frequency counters used in CMI calculations based on the 'i'th row of the data matrix.

    i

    the index for current data row

    f_C

    frequency table of class C

    f_X

    frequency table of feature X

    f_CX

    joint frequency table of C and X

    f_CXZ

    joint frequency table of C, X, and Z, where X and Z are features/columns

    Attributes
    protected
    Definition Classes
    TwoBAN_OS0BayesClassifier
  73. var vc: Array[Int]
    Attributes
    protected
  74. def vc_default: Array[Int]

    Return default values for binary input data (value count 'vc' set to 2).

    Return default values for binary input data (value count 'vc' set to 2).

    Definition Classes
    ClassifierInt
  75. def vc_fromData: Array[Int]

    Return value counts calculated from the input data.

    Return value counts calculated from the input data. May wish to call 'shiftToZero' before calling this method.

    Definition Classes
    ClassifierInt
  76. def vc_fromData2(rg: Range): Array[Int]

    Return value counts calculated from the input data.

    Return value counts calculated from the input data. May wish to call 'shiftToZero' before calling this method.

    rg

    the range of columns to be considered

    Definition Classes
    ClassifierInt
  77. val vcp1: Array[Int]
    Attributes
    protected
  78. val vcp2: Array[Int]
    Attributes
    protected
  79. final def wait(): Unit
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  80. final def wait(arg0: Long, arg1: Int): Unit
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  81. final def wait(arg0: Long): Unit
    Definition Classes
    AnyRef
    Annotations
    @native() @throws( ... )

Inherited from BayesClassifier

Inherited from BayesMetrics

Inherited from ClassifierInt

Inherited from Error

Inherited from Classifier

Inherited from AnyRef

Inherited from Any

Ungrouped