Serialized Form

Package smile.classification
- Class smile.classification.AdaBoost extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - k
```
int k
```
      The number of classes.
    - trees
```
DecisionTree[] trees
```
      Forest of decision trees.
    - alpha
```
double[] alpha
```
      The weight of each decision tree.
    - error
```
double[] error
```
      The weighted error of each decision tree during training.
    - importance
```
double[] importance
```
      Variable importance. Every time a split of a node is made on variable the (GINI, information gain, etc.) impurity criterion for the two descendent nodes is less than the parent node. Adding up the decreases for each individual variable over all trees in the forest gives a fast variable importance that is often very consistent with the permutation importance measure.
- Class smile.classification.DecisionTree extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - attributes
```
smile.data.Attribute[] attributes
```
      The attributes of independent variable.
    - importance
```
double[] importance
```
      Variable importance. Every time a split of a node is made on variable the (GINI, information gain, etc.) impurity criterion for the two descendent nodes is less than the parent node. Adding up the decreases for each individual variable over the tree gives a simple measure of variable importance.
    - root
```
smile.classification.DecisionTree.Node root
```
      The root of the regression tree
    - rule
```
DecisionTree.SplitRule rule
```
      The splitting rule.
    - k
```
int k
```
      The number of classes.
    - nodeSize
```
int nodeSize
```
      The minimum size of leaf nodes.
    - maxNodes
```
int maxNodes
```
      The maximum number of leaf nodes in the tree.
    - mtry
```
int mtry
```
      The number of input variables to be used to determine the decision at a node of the tree.
- Class smile.classification.FLD extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimensionality of data.
    - k
```
int k
```
      The number of classes.
    - mean
```
double[] mean
```
      Original common mean vector.
    - mu
```
smile.math.matrix.DenseMatrix mu
```
      Original class mean vectors.
    - scaling
```
smile.math.matrix.DenseMatrix scaling
```
      Project matrix.
    - smean
```
double[] smean
```
      Projected common mean vector.
    - smu
```
double[][] smu
```
      Projected class mean vectors.
- Class smile.classification.GradientTreeBoost extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - k
```
int k
```
      The number of classes.
    - trees
```
RegressionTree[] trees
```
      Forest of regression trees for binary classification.
    - forest
```
RegressionTree[][] forest
```
      Forest of regression trees for multi-class classification.
    - importance
```
double[] importance
```
      Variable importance. Every time a split of a node is made on variable the impurity criterion for the two descendent nodes is less than the parent node. Adding up the decreases for each individual variable over all trees in the forest gives a simple variable importance.
    - b
```
double b
```
      The intercept.
    - shrinkage
```
double shrinkage
```
      The shrinkage parameter in (0, 1] controls the learning rate of procedure.
    - maxNodes
```
int maxNodes
```
      The number of leaves in each tree.
    - ntrees
```
int ntrees
```
      The number of trees.
    - subsample
```
double subsample
```
      The sampling rate for stochastic tree boosting.
- Class smile.classification.KNN extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - knn
```
KNNSearch<K,V> knn
```
      The data structure for nearest neighbor search.
    - y
```
int[] y
```
      The labels of training samples.
    - k
```
int k
```
      The number of neighbors for decision.
    - c
```
int c
```
      The number of classes.
- Class smile.classification.LDA extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimensionality of data.
    - k
```
int k
```
      The number of classes.
    - ct
```
double[] ct
```
      Constant term of discriminant function of each class.
    - priori
```
double[] priori
```
      A priori probabilities of each class.
    - mu
```
double[][] mu
```
      Mean vectors of each class.
    - scaling
```
smile.math.matrix.DenseMatrix scaling
```
      Eigen vectors of common covariance matrix, which transforms observations to discriminant functions, normalized so that common covariance matrix is spherical.
    - eigen
```
double[] eigen
```
      Eigen values of common variance matrix.
- Class smile.classification.LogisticRegression extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimension of input space.
    - k
```
int k
```
      The number of classes.
    - L
```
double L
```
      The log-likelihood of learned model.
    - w
```
double[] w
```
      The linear weights for binary logistic regression.
    - W
```
double[][] W
```
      The linear weights for multi-class logistic regression.
    - lambda
```
double lambda
```
      Regularization factor.
    - eta
```
double eta
```
      learning rate for stochastic gradient descent.
- Class smile.classification.Maxent extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimension of input space.
    - k
```
int k
```
      The number of classes.
    - L
```
double L
```
      The log-likelihood of learned model.
    - w
```
double[] w
```
      The linear weights for binary logistic regression.
    - W
```
double[][] W
```
      The linear weights for multi-class logistic regression.
- Class smile.classification.NaiveBayes extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - model
```
NaiveBayes.Model model
```
      The generation model of naive Bayes.
    - k
```
int k
```
      The number of classes.
    - p
```
int p
```
      The number of independent variables.
    - priori
```
double[] priori
```
      The priori probability of each class.
    - prob
```
smile.stat.distribution.Distribution[][] prob
```
      The conditional distribution for general purpose naive Bayes classifier.
    - sigma
```
double sigma
```
      Amount of add-k smoothing of evidence. By default, we use add-one or Laplace smoothing, which simply adds one to each count to eliminate zeros. Add-one smoothing can be interpreted as a uniform prior (each term occurs once for each class) that is then updated as evidence from the training data comes in.
    - predefinedPriori
```
boolean predefinedPriori
```
      True if the priori probabilities are pre defined by user.
    - n
```
int n
```
      The total number of instances.
    - nc
```
int[] nc
```
      The number of instances in each class.
    - nt
```
int[] nt
```
      The number of terms of each class.
    - ntc
```
int[][] ntc
```
      The number of each term in each class.
    - condprob
```
double[][] condprob
```
      The log conditional probabilities for document classification.
- Class smile.classification.NeuralNetwork extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - errorFunction
```
NeuralNetwork.ErrorFunction errorFunction
```
      The type of error function of network.
    - activationFunction
```
NeuralNetwork.ActivationFunction activationFunction
```
      The type of activation function in output layer.
    - p
```
int p
```
      The dimensionality of data.
    - k
```
int k
```
      The number of classes.
    - net
```
smile.classification.NeuralNetwork.Layer[] net
```
      layers of this net
    - inputLayer
```
smile.classification.NeuralNetwork.Layer inputLayer
```
      input layer
    - outputLayer
```
smile.classification.NeuralNetwork.Layer outputLayer
```
      output layer
    - eta
```
double eta
```
      learning rate
    - alpha
```
double alpha
```
      momentum factor
    - lambda
```
double lambda
```
      weight decay factor, which is also a regularization term.
    - target
```
double[] target
```
      The buffer to store target value of training instance.
- Class smile.classification.PlattScaling extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - alpha
```
double alpha
```
      The scalar parameters to be learned by the algorithm.
    - beta
```
double beta
```
- Class smile.classification.QDA extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimensionality of data.
    - k
```
int k
```
      The number of classes.
    - ct
```
double[] ct
```
      Constant term of discriminant function of each class.
    - priori
```
double[] priori
```
      A priori probabilities of each class.
    - mu
```
double[][] mu
```
      Mean vectors of each class.
    - scaling
```
smile.math.matrix.DenseMatrix[] scaling
```
      Eigen vectors of each covariance matrix, which transforms observations to discriminant functions, normalized so that within groups covariance matrix is spherical.
    - ev
```
double[][] ev
```
      Eigen values of each covariance matrix.
- Class smile.classification.RandomForest extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - trees
```
java.util.List<E> trees
```
      Forest of decision trees. The second value is the accuracy of tree on the OOB samples, which can be used a weight when aggregating tree votes.
    - k
```
int k
```
      The number of classes.
    - error
```
double error
```
      Out-of-bag estimation of error rate, which is quite accurate given that enough trees have been grown (otherwise the OOB estimate can bias upward).
    - importance
```
double[] importance
```
      Variable importance. Every time a split of a node is made on variable the (GINI, information gain, etc.) impurity criterion for the two descendent nodes is less than the parent node. Adding up the decreases for each individual variable over all trees in the forest gives a fast variable importance that is often very consistent with the permutation importance measure.
- Class smile.classification.RBFNetwork extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - k
```
int k
```
      The number of classes.
    - centers
```
java.lang.Object[] centers
```
      The centers of RBF functions.
    - w
```
smile.math.matrix.DenseMatrix w
```
      The linear weights.
    - distance
```
smile.math.distance.Metric<T> distance
```
      The distance metric functor.
    - rbf
```
smile.math.rbf.RadialBasisFunction[] rbf
```
      The radial basis function.
    - normalized
```
boolean normalized
```
      True to fit a normalized RBF network.
- Class smile.classification.RDA extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimensionality of data.
    - k
```
int k
```
      The number of classes.
    - ct
```
double[] ct
```
      Constant term of discriminant function of each class.
    - priori
```
double[] priori
```
      A priori probabilities of each class.
    - mu
```
double[][] mu
```
      Mean vectors of each class.
    - scaling
```
smile.math.matrix.DenseMatrix[] scaling
```
      Eigen vectors of each covariance matrix, which transforms observations to discriminant functions, normalized so that within groups covariance matrix is spherical.
    - ev
```
double[][] ev
```
      Eigen values of each covariance matrix.
- Class smile.classification.SVM extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - svm
```
smile.classification.SVM.LASVM svm
```
      Learned two-class support vector machine.
    - svms
```
java.util.List<E> svms
```
      Learned multi-class support vector machines.
    - kernel
```
smile.math.kernel.MercerKernel<T> kernel
```
      The kernel function.
    - p
```
int p
```
      The dimensionality of instances. Useful for sparse arrays.
    - k
```
int k
```
      The number of classes;
    - strategy
```
SVM.Multiclass strategy
```
      The strategy for multi-class classification.
    - wi
```
double[] wi
```
      The class weight.
    - tol
```
double tol
```
      The tolerance of convergence test.
- Class smile.classification.SVM.SupportVector extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - x
```
java.lang.Object x
```
      Support vector.
    - y
```
int y
```
      Support vector label.
    - alpha
```
double alpha
```
      Lagrangian multiplier of support vector.
Package smile.clustering
- Class smile.clustering.BIRCH extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - B
```
int B
```
      Branching factor. Maximum number of children nodes.
    - T
```
double T
```
      Maximum radius of a sub-cluster.
    - d
```
int d
```
      The dimensionality of data.
    - root
```
smile.clustering.BIRCH.Node root
```
      The root of CF tree.
    - centroids
```
double[][] centroids
```
      Leaves of CF tree as representatives of all data points.
- Class smile.clustering.CLARANS extends PartitionClustering<T> implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - distortion
```
double distortion
```
      The total distortion.
    - distance
```
smile.math.distance.Distance<T> distance
```
      The distance measure for calculation of distortion.
    - numLocal
```
int numLocal
```
      The number of local minima to search for.
    - maxNeighbor
```
int maxNeighbor
```
      The maximum number of neighbors examined during a search of local minima.
    - medoids
```
java.lang.Object[] medoids
```
      The medoids of each cluster.
- Class smile.clustering.DBSCAN extends PartitionClustering<T> implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - minPts
```
double minPts
```
      The minimum number of points required to form a cluster
    - radius
```
double radius
```
      The range of neighborhood.
    - nns
```
RNNSearch<K,V> nns
```
      Data structure for neighborhood search.
- Class smile.clustering.DENCLUE extends PartitionClustering<double[]> implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - eps
```
double eps
```
      The epsilon of finishing density attractor searching.
    - sigma
```
double sigma
```
      The smooth parameter in the Gaussian kernel. The user can choose sigma such that number of density attractors is constant for a long interval of sigma.
    - gamma
```
double gamma
```
      The smooth parameter in the Gaussian kernel. It is -0.5 / (sigma * sigma).
    - attractors
```
double[][] attractors
```
      The density attractor of each cluster.
    - radius
```
double[] radius
```
      The radius of density attractor.
    - samples
```
double[][] samples
```
      The samples decided by NeuralGas used in the iterations of hill climbing.
- Class smile.clustering.DeterministicAnnealing extends KMeans implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - alpha
```
double alpha
```
      Annealing parameter in (0, 1).
- Class smile.clustering.GMeans extends KMeans implements Serializable
  
  serialVersionUID:
  
  1L
- Class smile.clustering.KMeans extends PartitionClustering<double[]> implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - distortion
```
double distortion
```
      The total distortion.
    - centroids
```
double[][] centroids
```
      The centroids of each cluster.
- Class smile.clustering.MEC extends PartitionClustering<T> implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - radius
```
double radius
```
      The range of neighborhood.
    - entropy
```
double entropy
```
      The conditional entropy as the objective function.
    - nns
```
RNNSearch<K,V> nns
```
      The neighborhood search data structure.
- Class smile.clustering.PartitionClustering extends java.lang.Object implements Serializable
  - Serialized Fields
    - k
```
int k
```
      The number of clusters.
    - y
```
int[] y
```
      The cluster labels of data.
    - size
```
int[] size
```
      The number of samples in each cluster.
- Class smile.clustering.SIB extends PartitionClustering<double[]> implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - distortion
```
double distortion
```
      The total distortion.
    - centroids
```
double[][] centroids
```
      The centroids of each cluster.
- Class smile.clustering.XMeans extends KMeans implements Serializable
  
  serialVersionUID:
  
  1L
Package smile.feature
- Class smile.feature.FeatureTransform extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - copy
```
boolean copy
```
      If false, try to avoid a copy and do inplace transformation instead.
- Class smile.feature.MaxAbsScaler extends FeatureTransform implements Serializable
  - Serialized Fields
    - scale
```
double[] scale
```
      Scaling factor.
- Class smile.feature.Normalizer extends FeatureTransform implements Serializable
  - Serialized Fields
    - norm
```
Normalizer.Norm norm
```
      The type of norm .
- Class smile.feature.RobustStandardizer extends Standardizer implements Serializable
- Class smile.feature.Scaler extends FeatureTransform implements Serializable
  - Serialized Fields
    - lo
```
double[] lo
```
      Lower bound.
    - hi
```
double[] hi
```
      Upper bound.
- Class smile.feature.Standardizer extends FeatureTransform implements Serializable
  - Serialized Fields
    - mu
```
double[] mu
```
      Mean or median.
    - std
```
double[] std
```
      Standard deviation or IQR.
- Class smile.feature.WinsorScaler extends Scaler implements Serializable
  - Serialized Fields
    - lower
```
double lower
```
      Lower limit.
    - upper
```
double upper
```
      Upper limit.
Package smile.imputation
- Class smile.imputation.MissingValueImputationException extends java.lang.Exception implements Serializable
  
  serialVersionUID:
  
  -4517157687017284522L
Package smile.neighbor
- Class smile.neighbor.BKTree extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - root
```
smile.neighbor.BKTree.Node root
```
      The root root of BK-tree.
    - distance
```
smile.math.distance.Metric<T> distance
```
      The distance metric. Note that the metric must be a discrete distance, e.g. edit distance, Hamming distance, Lee distance, Jaccard distance, and taxonomic distance, etc.
    - count
```
int count
```
      The number of nodes in the tree.
    - identicalExcluded
```
boolean identicalExcluded
```
      Whether to exclude query object self from the neighborhood.
- Class smile.neighbor.CoverTree extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - data
```
java.lang.Object[] data
```
      The dataset to build the cover tree.
    - distance
```
smile.math.distance.Metric<T> distance
```
      The distance/metric function for nearest neighbor search.
    - root
```
smile.neighbor.CoverTree.Node root
```
      The root node.
    - base
```
double base
```
      The base of our expansion constant. In other words the 2 in 2^i used in covering tree and separation invariants of a cover tree. In paper it's suggested the separation invariant is relaxed in batch construction.
    - invLogBase
```
double invLogBase
```
      if we have base 2 then this can be viewed as 1/ln(2), which can be used later on to do invLogBase*ln(d) instead of ln(d)/ln(2), to get log2(d), in getScale method.
    - identicalExcluded
```
boolean identicalExcluded
```
      Whether to exclude query object self from the neighborhood.
- Class smile.neighbor.KDTree extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - keys
```
double[][] keys
```
      The keys of data objects.
    - data
```
java.lang.Object[] data
```
      The data objects.
    - root
```
smile.neighbor.KDTree.Node root
```
      The root node of KD-Tree.
    - index
```
int[] index
```
      The index of objects in each nodes.
    - identicalExcluded
```
boolean identicalExcluded
```
      Whether to exclude query object self from the neighborhood.
- Class smile.neighbor.LSH extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - P
```
int P
```
      The prime number in universal bucket hashing.
    - MAX_HASH_RND
```
int MAX_HASH_RND
```
      The range of universal hashing random integers [0, 2^29).
    - keys
```
java.util.ArrayList<E> keys
```
      The keys of data objects.
    - data
```
java.util.ArrayList<E> data
```
      The data objects.
    - hash
```
java.util.List<E> hash
```
      Hash functions.
    - H
```
int H
```
      The size of hash table.
    - d
```
int d
```
      The dimensionality of data.
    - L
```
int L
```
      The number of hash tables.
    - k
```
int k
```
      The number of random projections per hash value.
    - w
```
double w
```
      The width of projection. The hash function is defined as floor((a * x + b) / w). The value of w determines the bucket interval.
    - r1
```
int[] r1
```
      The random integer used for universal bucket hashing.
    - r2
```
int[] r2
```
      The random integer used for universal hashing for control values.
    - identicalExcluded
```
boolean identicalExcluded
```
      Whether to exclude query object self from the neighborhood.
- Class smile.neighbor.MPLSH extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - keys
```
java.util.ArrayList<E> keys
```
      The keys of data objects.
    - data
```
java.util.ArrayList<E> data
```
      The data objects.
    - hash
```
java.util.List<E> hash
```
      Hash functions.
    - H
```
int H
```
      The size of hash table.
    - d
```
int d
```
      The dimensionality of data.
    - L
```
int L
```
      The number of hash tables.
    - k
```
int k
```
      The number of random projections per hash value.
    - r
```
double r
```
      The width of projection. The hash function is defined as floor((a * x + b) / w). The value of w determines the bucket interval.
    - c
```
int[] c
```
      The random integer used for universal bucket hashing.
    - P
```
int P
```
      The prime number in universal bucket hashing.
    - identicalExcluded
```
boolean identicalExcluded
```
      Whether to exclude query object self from the neighborhood.
    - model
```
java.util.List<E> model
```
      Pre-computed posteriori lookup table to generate multiple probes.
- Class smile.neighbor.SNLSH extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - bandSize
```
int bandSize
```
    - mask
```
long mask
```
    - bands
```
smile.neighbor.SNLSH.Band[] bands
```
      Signature fractions
    - data
```
java.util.List<E> data
```
      The data objects.
    - keys
```
java.util.List<E> keys
```
      The keys of data objects.
    - signs
```
java.util.List<E> signs
```
      signatures generated by simhash
    - identicalExcluded
```
boolean identicalExcluded
```
      Whether to exclude query object self from the neighborhood.
Package smile.projection
- Class smile.projection.GHA extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimension of feature space.
    - n
```
int n
```
      The dimension of input space.
    - r
```
double r
```
      The learning rate;
    - projection
```
smile.math.matrix.DenseMatrix projection
```
      Projection matrix.
    - y
```
double[] y
```
      Workspace for W * x.
    - wy
```
double[] wy
```
      Workspace for W' * y.
- Class smile.projection.KPCA extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimension of feature space.
    - data
```
java.lang.Object[] data
```
      Learning data.
    - kernel
```
smile.math.kernel.MercerKernel<T> kernel
```
      Mercer kernel.
    - mean
```
double[] mean
```
      The row mean of kernel matrix.
    - mu
```
double mu
```
      The mean of kernel matrix.
    - latent
```
double[] latent
```
      Eigenvalues of kernel principal components.
    - projection
```
smile.math.matrix.DenseMatrix projection
```
      Projection matrix.
    - coordinates
```
double[][] coordinates
```
      The coordinates of projected training data.
- Class smile.projection.PCA extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimension of feature space.
    - n
```
int n
```
      The dimension of input space.
    - mu
```
double[] mu
```
      The sample mean.
    - pmu
```
double[] pmu
```
      The projected sample mean.
    - eigvectors
```
smile.math.matrix.DenseMatrix eigvectors
```
      The matrix of variable loadings, whose columns contain the eigenvectors.
    - eigvalues
```
double[] eigvalues
```
      Eigenvalues of principal components.
    - proportion
```
double[] proportion
```
      The proportion of variance contained in each principal component.
    - cumulativeProportion
```
double[] cumulativeProportion
```
      The cumulative proportion of variance contained in principal components.
    - projection
```
smile.math.matrix.DenseMatrix projection
```
      Projection matrix.
- Class smile.projection.PPCA extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - mu
```
double[] mu
```
      The sample mean.
    - pmu
```
double[] pmu
```
      The projected sample mean.
    - noise
```
double noise
```
      The variance of noise part.
    - loading
```
smile.math.matrix.DenseMatrix loading
```
      Loading matrix.
    - projection
```
smile.math.matrix.DenseMatrix projection
```
      Projection matrix.
- Class smile.projection.RandomProjection extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimension of feature space.
    - n
```
int n
```
      The dimension of input space.
    - projection
```
smile.math.matrix.DenseMatrix projection
```
      Projection matrix.
Package smile.regression
- Class smile.regression.ElasticNet extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - lambda1
```
double lambda1
```
      parameter for L1 regularization
    - lambda2
```
double lambda2
```
      parameter for L2 regularization
    - p
```
int p
```
      The dimensionality.
    - w
```
double[] w
```
      corrected coefficients
    - b
```
double b
```
      corrected intercept
    - lasso
```
LASSO lasso
```
      reduced to lasso problem
    - c
```
double c
```
      scaling calculated from lambda2
- Class smile.regression.GaussianProcessRegression extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - knots
```
java.lang.Object[] knots
```
      The control points in the regression.
    - w
```
double[] w
```
      The linear weights.
    - kernel
```
smile.math.kernel.MercerKernel<T> kernel
```
      The distance functor.
    - lambda
```
double lambda
```
      The shrinkage/regularization parameter.
- Class smile.regression.GradientTreeBoost extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - trees
```
RegressionTree[] trees
```
      Forest of regression trees.
    - b
```
double b
```
      The intercept.
    - importance
```
double[] importance
```
      Variable importance. Every time a split of a node is made on variable the impurity criterion for the two descendent nodes is less than the parent node. Adding up the decreases for each individual variable over all trees in the forest gives a simple variable importance.
    - loss
```
GradientTreeBoost.Loss loss
```
      Loss function.
    - shrinkage
```
double shrinkage
```
      The shrinkage parameter in (0, 1] controls the learning rate of procedure.
    - maxNodes
```
int maxNodes
```
      The number of leaves in each tree.
    - ntrees
```
int ntrees
```
      The number of trees.
    - f
```
double f
```
      The sampling rate for stochastic tree boosting.
- Class smile.regression.LASSO extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimensionality.
    - lambda
```
double lambda
```
      The shrinkage/regularization parameter.
    - b
```
double b
```
      The intercept.
    - w
```
double[] w
```
      The linear coefficients.
    - ym
```
double ym
```
      The mean of response variable.
    - center
```
double[] center
```
      The center of input vector. The input vector should be centered before prediction.
    - scale
```
double[] scale
```
      Scaling factor of input vector.
    - residuals
```
double[] residuals
```
      The residuals, that is response minus fitted values.
    - RSS
```
double RSS
```
      Residual sum of squares.
    - error
```
double error
```
      Residual standard error.
    - df
```
int df
```
      The degree-of-freedom of residual standard error.
    - RSquared
```
double RSquared
```
      R². R² is a statistic that will give some information about the goodness of fit of a model. In regression, the R² coefficient of determination is a statistical measure of how well the regression line approximates the real data points. An R² of 1.0 indicates that the regression line perfectly fits the data.
      In the case of ordinary least-squares regression, R² increases as we increase the number of variables in the model (R² will not decrease). This illustrates a drawback to one possible use of R², where one might try to include more variables in the model until "there is no more improvement". This leads to the alternative approach of looking at the adjusted R².
    - adjustedRSquared
```
double adjustedRSquared
```
      Adjusted R². The adjusted R² has almost same explanation as R² but it penalizes the statistic as extra variables are included in the model.
    - F
```
double F
```
      The F-statistic of the goodness-of-fit of the model.
    - pvalue
```
double pvalue
```
      The p-value of the goodness-of-fit test of the model.
- Class smile.regression.NeuralNetwork extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - activationFunction
```
NeuralNetwork.ActivationFunction activationFunction
```
      The type of activation function in output layer.
    - p
```
int p
```
      The dimensionality of data.
    - net
```
smile.regression.NeuralNetwork.Layer[] net
```
      layers of this net
    - inputLayer
```
smile.regression.NeuralNetwork.Layer inputLayer
```
      input layer
    - outputLayer
```
smile.regression.NeuralNetwork.Layer outputLayer
```
      output layer
    - eta
```
double eta
```
      learning rate
    - alpha
```
double alpha
```
      momentum factor
    - lambda
```
double lambda
```
      weight decay factor, which is also a regularization term.
- Class smile.regression.OLS extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimensionality.
    - b
```
double b
```
      The intercept.
    - w
```
double[] w
```
      The linear weights.
    - attributes
```
smile.data.Attribute[] attributes
```
      The variable attributes.
    - coefficients
```
double[][] coefficients
```
      The coefficients, their standard errors, t-scores, and p-values.
    - fittedValues
```
double[] fittedValues
```
      The fitted values.
    - residuals
```
double[] residuals
```
      The residuals, that is response minus fitted values.
    - RSS
```
double RSS
```
      Residual sum of squares.
    - error
```
double error
```
      Residual standard error.
    - df
```
int df
```
      The degree-of-freedom of residual standard error.
    - RSquared
```
double RSquared
```
      R². R² is a statistic that will give some information about the goodness of fit of a model. In regression, the R² coefficient of determination is a statistical measure of how well the regression line approximates the real data points. An R² of 1.0 indicates that the regression line perfectly fits the data.
      In the case of ordinary least-squares regression, R² increases as we increase the number of variables in the model (R² will not decrease). This illustrates a drawback to one possible use of R², where one might try to include more variables in the model until "there is no more improvement". This leads to the alternative approach of looking at the adjusted R².
    - adjustedRSquared
```
double adjustedRSquared
```
      Adjusted R². The adjusted R² has almost same explanation as R² but it penalizes the statistic as extra variables are included in the model.
    - F
```
double F
```
      The F-statistic of the goodness-of-fit of the model.
    - pvalue
```
double pvalue
```
      The p-value of the goodness-of-fit test of the model.
- Class smile.regression.RandomForest extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - trees
```
java.util.List<E> trees
```
      Forest of regression trees.
    - error
```
double error
```
      Out-of-bag estimation of RMSE, which is quite accurate given that enough trees have been grown (otherwise the OOB estimate can bias upward).
    - importance
```
double[] importance
```
      Variable importance. Every time a split of a node is made on variable the impurity criterion for the two descendent nodes is less than the parent node. Adding up the decreases for each individual variable over all trees in the forest gives a fast variable importance that is often very consistent with the permutation importance measure.
    - monotonicRegression
```
double[] monotonicRegression
```
      Values between [-1, 1] that represents monotonic regression coefficient for each attribute. It can be used to enforce model to keep monotonic relationship between target and the attribute. Positive value enforce target to be positively correlated with this feature. Positive value enforce target to be negatively correlated with this feature. Zero value turns off monotonic regression.
- Class smile.regression.RBFNetwork extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - centers
```
java.lang.Object[] centers
```
      The centers of RBF functions.
    - w
```
double[] w
```
      The linear weights.
    - distance
```
smile.math.distance.Metric<T> distance
```
      The distance functor.
    - rbf
```
smile.math.rbf.RadialBasisFunction[] rbf
```
      The radial basis functions.
    - normalized
```
boolean normalized
```
      True to fit a normalized RBF network.
- Class smile.regression.RegressionTree extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - attributes
```
smile.data.Attribute[] attributes
```
      The attributes of independent variable.
    - importance
```
double[] importance
```
      Variable importance. Every time a split of a node is made on variable the impurity criterion for the two descendent nodes is less than the parent node. Adding up the decreases for each individual variable over the tree gives a simple measure of variable importance.
    - monotonicRegression
```
double[] monotonicRegression
```
      Values between [-1, 1] that represents monotonic regression coefficient for each attribute. It can be used to enforce model to keep monotonic relationship between target and the attribute. Positive value enforce target to be positively correlated with this feature. Positive value enforce target to be negatively correlated with this feature. Zero value turns off monotonic regression.
    - root
```
smile.regression.RegressionTree.Node root
```
      The root of the regression tree
    - nodeSize
```
int nodeSize
```
      The number of instances in a node below which the tree will not split, setting nodeSize = 5 generally gives good results.
    - maxNodes
```
int maxNodes
```
      The maximum number of leaf nodes in the tree.
    - mtry
```
int mtry
```
      The number of input variables to be used to determine the decision at a node of the tree.
    - numFeatures
```
int numFeatures
```
      The number of binary features.
- Class smile.regression.RidgeRegression extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimensionality.
    - lambda
```
double lambda
```
      The shrinkage/regularization parameter.
    - b
```
double b
```
      The centered intercept.
    - w
```
double[] w
```
      The scaled linear coefficients.
    - ym
```
double ym
```
      The mean of response variable.
    - center
```
double[] center
```
      The center of input vector. The input vector should be centered before prediction.
    - scale
```
double[] scale
```
      Scaling factor of input vector.
    - coefficients
```
double[][] coefficients
```
      The coefficients, their standard errors, t-scores, and p-values.
    - residuals
```
double[] residuals
```
      The residuals, that is response minus fitted values.
    - RSS
```
double RSS
```
      Residual sum of squares.
    - error
```
double error
```
      Residual standard error.
    - df
```
int df
```
      The degree-of-freedom of residual standard error.
    - RSquared
```
double RSquared
```
      R². R² is a statistic that will give some information about the goodness of fit of a model. In regression, the R² coefficient of determination is a statistical measure of how well the regression line approximates the real data points. An R² of 1.0 indicates that the regression line perfectly fits the data.
      In the case of ordinary least-squares regression, R² increases as we increase the number of variables in the model (R² will not decrease). This illustrates a drawback to one possible use of R², where one might try to include more variables in the model until "there is no more improvement". This leads to the alternative approach of looking at the adjusted R².
    - adjustedRSquared
```
double adjustedRSquared
```
      Adjusted R². The adjusted R² has almost same explanation as R² but it penalizes the statistic as extra variables are included in the model.
    - F
```
double F
```
      The F-statistic of the goodness-of-fit of the model.
    - pvalue
```
double pvalue
```
      The p-value of the goodness-of-fit test of the model.
- Class smile.regression.RLS extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - p
```
int p
```
      The dimensionality.
    - w
```
double[] w
```
      The coefficients with intercept.
    - lambda
```
double lambda
```
      The forgetting factor in (0, 1]. Values closer to 1 will have longer memory and values closer to 0 will be have shorter memory.
    - V
```
smile.math.matrix.DenseMatrix V
```
      First initialized to the matrix (X^TX)^-1, it is updated with each new learning instance.
    - x1
```
double[] x1
```
      A single learning instance X, padded with 1 for intercept.
    - Vx
```
double[] Vx
```
      A temporary array used in computing V * X .
- Class smile.regression.SVR extends java.lang.Object implements Serializable
  
  serialVersionUID:
  
  1L
  - Serialized Fields
    - kernel
```
smile.math.kernel.MercerKernel<T> kernel
```
      The kernel function.
    - C
```
double C
```
      The soft margin penalty parameter.
    - eps
```
double eps
```
      The loss function error threshold.
    - tol
```
double tol
```
      The tolerance of convergence test.
    - sv
```
java.util.List<E> sv
```
      Support vectors.
    - b
```
double b
```
      Threshold of decision function.
    - nsv
```
int nsv
```
      The number of support vectors.
    - nbsv
```
int nbsv
```
      The number of bounded support vectors.
Package smile.sequence
- Class smile.sequence.CRF extends java.lang.Object implements Serializable
  - Serialized Fields
    - numClasses
```
int numClasses
```
      The number of classes.
    - numFeatures
```
int numFeatures
```
      The number of sparse binary features.
    - potentials
```
smile.sequence.CRF.TreePotentialFunction[] potentials
```
      The potential functions for each class.
    - viterbi
```
boolean viterbi
```
      True if using Viterbi algorithm for sequence labeling.
Package smile.taxonomy
- Class smile.taxonomy.TaxonomicDistance extends java.lang.Object implements Serializable
  - Serialized Fields
    - taxonomy
```
Taxonomy taxonomy
```
      The taxonomy basis to calculate the distance.
Package smile.vq
- Class smile.vq.GrowingNeuralGas extends java.lang.Object implements Serializable
  - Serialized Fields
    - d
```
int d
```
      The dimensionality of signals.
    - n
```
int n
```
      The number of signals processed so far.
    - m
```
int m
```
      The number of neurons created so far.
    - epsBest
```
double epsBest
```
      The fraction to update nearest neuron.
    - epsNeighbor
```
double epsNeighbor
```
      The fraction to update neighbors of nearest neuron.
    - maxEdgeAge
```
int maxEdgeAge
```
      The maximum age of edges.
    - lambda
```
int lambda
```
      If the number of input signals so far is an integer multiple of lambda, insert a new neuron.
    - alpha
```
double alpha
```
      Decrease error variables by multiplying them with alpha during inserting a new neuron.
    - beta
```
double beta
```
      Decrease all error variables by multiply them with de.
    - nodes
```
java.util.LinkedList<E> nodes
```
      Neurons in the neural network.
    - y
```
int[] y
```
      Cluster labels of neurons.
- Class smile.vq.NeuralGas extends PartitionClustering<double[]> implements Serializable
  - Serialized Fields
    - distortion
```
double distortion
```
      The total distortion.
    - centroids
```
double[][] centroids
```
      The centroids of each cluster.
- Class smile.vq.NeuralMap extends java.lang.Object implements Serializable
  - Serialized Fields
    - d
```
int d
```
      The dimensionality of signals.
    - r
```
double r
```
      The distance radius to activate a neuron for a given signal.
    - epsBest
```
double epsBest
```
      The fraction to update nearest neuron.
    - epsNeighbor
```
double epsNeighbor
```
      The fraction to update neighbors of nearest neuron.
    - lsh
```
smile.vq.NeuralMap.LSH lsh
```
      Neurons in the neural network.
    - neurons
```
java.util.List<E> neurons
```
      The list of neurons.
- Class smile.vq.SOM extends java.lang.Object implements Serializable
  - Serialized Fields
    - width
```
int width
```
      The width of map.
    - height
```
int height
```
      The height of map.
    - d
```
int d
```
      The dimension of input space.
    - neurons
```
double[][][] neurons
```
      The SOM map of neurons.
    - bmu
```
int[][] bmu
```
      The best matched unit. This is n-by-2 matrix, of which each row is for each data point. The entry bmu[i][0] and bmu[i][1] are the row index and column index of the best matched unit for each sample, respectively.
    - voronoi
```
int[][] voronoi
```
      The number of samples in each units.
    - y
```
int[] y
```
      The cluster labels of neurons.
    - umatrix
```
double[][] umatrix
```
      Unified distance matrix, or u-matrix, is a popular method of displaying SOMs. The value of umatrix is the maximum of distances between a map unit to its neighbors.

Serialized Form

Package smile.classification

Class smile.classification.AdaBoost extends java.lang.Object implements Serializable

Serialized Fields

k

trees

alpha

error

importance

Class smile.classification.DecisionTree extends java.lang.Object implements Serializable

Serialized Fields

attributes

importance

root

rule

k

nodeSize

maxNodes

mtry

Class smile.classification.FLD extends java.lang.Object implements Serializable

Serialized Fields

p

k

mean

mu

scaling

smean

smu

Class smile.classification.GradientTreeBoost extends java.lang.Object implements Serializable

Serialized Fields

k

trees

forest

importance

b

shrinkage

maxNodes

ntrees

subsample

Class smile.classification.KNN extends java.lang.Object implements Serializable

Serialized Fields

knn

y

k

c

Class smile.classification.LDA extends java.lang.Object implements Serializable

Serialized Fields

p

k

ct

priori

mu

scaling

eigen

Class smile.classification.LogisticRegression extends java.lang.Object implements Serializable

Serialized Fields

p

k

L

w

W

lambda

eta

Class smile.classification.Maxent extends java.lang.Object implements Serializable

Serialized Fields

p

k

L

w

W

Class smile.classification.NaiveBayes extends java.lang.Object implements Serializable

Serialized Fields

model

k

p

priori

prob

sigma

predefinedPriori

n