oracle.dmt.jdm.algorithm.tree
Class OraTreeSettings

java.lang.Object
oracle.dmt.jdm.OraDMObject
oracle.dmt.jdm.base.OraAlgorithmSettings
oracle.dmt.jdm.supervised.OraSupervisedAlgorithmSettings
oracle.dmt.jdm.algorithm.tree.OraTreeSettings

All Implemented Interfaces:

javax.datamining.base.AlgorithmSettings, oracle.dmt.jdm.OraPLSQLConstants, javax.datamining.supervised.SupervisedAlgorithmSettings, javax.datamining.algorithm.tree.TreeSettings

public class OraTreeSettings

extends oracle.dmt.jdm.supervised.OraSupervisedAlgorithmSettings

implements javax.datamining.algorithm.tree.TreeSettings

OraTreeSettings defines oracle specific decision tree algorithm settings i.e., minimum decrease in impurity.

• Minimum decrease in impurity required to justify splitting a node. This is a termination criterion. If no candidate split can be found with a decrease greater than the minimum, then the build process halts. The minimum impurity must be non-negative. The minimum impurity can be specified in terms of the number of records at a node or the ratio of the records at a node to the entire data.

Field Summary

Fields inherited from interface oracle.dmt.jdm.OraPLSQLConstants

abns_max_build_minutes, abns_max_nb_predictors, abns_max_predictors, abns_model_type, abns_multi_feature, abns_naive_bayes, abns_single_feature, algo_adaptive_bayes_network, algo_ai_mdl, algo_ai_mdl2, algo_apriori_association_rules, algo_decision_tree, algo_kmeans, algo_naive_bayes, algo_name, algo_nonnegative_matrix_factor, algo_ocluster, algo_predictor_variance, algo_support_vector_machines, apply_cost_content, apply_nodeid_content, apply_pred_value_content, apply_probability_content, asso_max_rule_length, asso_min_confidence, asso_min_support, association, association_in_model, attribute_importance, clas_cost_table_name, clas_priors_table_name, classification, clus_num_clusters, clustering, feat_num_features, feature_extraction, kmns_block_growth, kmns_conv_tolerance, kmns_cosine, kmns_distance, kmns_euclidean, kmns_fast_cosine, kmns_iterations, kmns_min_pct_attr_support, kmns_num_bins, kmns_size, kmns_split_criterion, kmns_variance, nabs_pairwise_threshold, nabs_singleton_threshold, nmfs_conv_tolerance, nmfs_num_iterations, nmfs_random_seed, ocluster_max_buffer, ocluster_sensitivity, operator_equal, operator_equal_v, operator_greater_or_equal, operator_greater_or_equal_v, operator_greater_than, operator_greater_than_v, operator_in, operator_in_v, operator_less_or_equal, operator_less_or_equal_v, operator_less_than, operator_less_than_v, operator_not_equal, operator_not_equal_v, operator_not_in, operator_not_in_v, oracle_char_type, oracle_dm_nested_categoricals, oracle_dm_nested_numericals, oracle_float_type, oracle_number_type, oracle_varchar2_type, regression, svms_active_learning, svms_al_disable, svms_al_enable, svms_complexity_factor, svms_conv_tolerance, svms_epsilon, svms_gaussian, svms_kernel_cache_size, svms_kernel_function, svms_linear, svms_outlier_rate, svms_std_dev, tree_impurity_entropy, tree_impurity_gini, tree_impurity_metric, tree_impurity_metric_default, tree_term_max_depth, tree_term_max_depth_default, tree_term_max_depth_max, tree_term_max_depth_min, tree_term_max_surrogates_max, tree_term_max_surrogates_min, tree_term_minpct_node, tree_term_minpct_node_default, tree_term_minpct_node_max, tree_term_minpct_split, tree_term_minpct_split_default, tree_term_minpct_split_max, tree_term_minrec_node, tree_term_minrec_node_default, tree_term_minrec_split, tree_term_minrec_split_default

Method Summary

double	getMinNodeSizeForSplit(javax.datamining.SizeUnit unit) Oracle extension.
void	setMinNodeSizeForSplit(double minNodeSize, javax.datamining.SizeUnit unit) Oracle extension.

Methods inherited from class oracle.dmt.jdm.base.OraAlgorithmSettings

getMiningAlgorithm, verify

Methods inherited from class oracle.dmt.jdm.OraDMObject

createException, createException, createRuntimeException, createRuntimeException, getLocalizedMessage, isConnectionOpen, logInfo, logSevere, logTrace, logTrace, unsupported, unsupported

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface javax.datamining.base.AlgorithmSettings

getMiningAlgorithm, verify

Method Detail

getMinNodeSizeForSplit

public double getMinNodeSizeForSplit(javax.datamining.SizeUnit unit)

Oracle extension. Returns the minimum node size for split. No split is made if the node size is less than this minimum.

Parameters:

unit - The size unit of the minimum node size for split.

Returns:

double

setMinNodeSizeForSplit

public void setMinNodeSizeForSplit(double minNodeSize,
                                   javax.datamining.SizeUnit unit)

Oracle extension. Sets the minimum node size for split. This is a termination criterion. No split is made if the node size is less than this minimum. The minimum node size for split must be positive. The minimum node size for split can be specified in terms of the number of records at a node or the ratio of the records at a node to the entire data. (Oracle extension)

Parameters:

minNodeSize - The minimum node size for split.

unit - The size unit of minNodeSize

Returns:

void