ClassificationBaggedEnsemble class
Superclasses:
Classification ensemble grown by resampling
描述
ClassificationBaggedEnsemblecombines a set of trained weak learner models and data on which these learners were trained. It can predict ensemble response for new data by aggregating predictions from its weak learners.
Construction
Create a bagged classification ensemble object usingfitcensemble.
Properties
|
Categorical predictor indices, specified as a vector of positive integers. |
|
元素列表 |
|
角色向量描述了如何 |
|
Expanded predictor names, stored as a cell array of character vectors. If the model uses encoding for categorical variables, then |
|
Numeric array of fit information. The |
|
角色向量描述了 |
|
Numeric scalar between |
|
描述的cross-validation optimization of hyperparameters, stored as a
|
|
Character vector describing the method that creates |
|
Parameters used in training |
|
Number of trained weak learners in |
|
预测变量的单元格数字，按照它们出现的顺序 |
|
Character vector describing the reason |
|
Logical value indicating if the ensemble was trained with replacement ( |
|
Character vector with the name of the response variable |
|
用于转换分数的功能句柄或代表内置转换函数的字符向量。 Add or change a ens.scoretransform ='功能' 或者 ens.scoretransform = @功能 |
|
受过训练的学习者，紧凑的分类模型的单元格数组。 |
|
较弱学习者的训练重量的数字向量 |
|
逻辑矩阵的大小 |
|
缩放 |
|
Matrix of predictor values that trained the ensemble. Each column of |
|
一个分类数组，字符向量的单元格数组，字符阵列，逻辑向量或数字向量与行相同数量的行 |
方法
| OOBEDGE | Out-of-bag classification edge |
| oobLoss | Out-of-bag classification error |
| OOBMARGIN | Out-of-bag classification margins |
| oobPermutedPredictorImportance | 预测指标通过对分类树的随机森林的排列外预测观测的置换来预测重要性估计 |
| oobPredict | Predict out-of-bag response of ensemble |
Inherited Methods
| 袖珍的 | 紧凑型分类合奏 |
| crossval | Cross validate ensemble |
| resubEdge | Classification edge by resubstitution |
| 重新公开 | 分类错误通过重述 |
| repubmargin | 通过重新确定的分类利润率 |
| resubPredict | 通过重新结算预测合奏的反应 |
| resume | 简历培训合奏 |
| 比较 | 使用新数据比较两个分类模型的精度 |
| 边缘 | Classification edge |
| loss | 分类错误 |
| margin | 分类边缘 |
| 预测 | Predict labels using ensemble of classification models |
| 预测或者Importance | 预测指标的重要性 |
| 删除者 | Remove members of compact classification ensemble |
Copy Semantics
价值。要了解价值类别如何影响复制操作，请参见复制对象（MATLAB）。
例子
Train Bagged Ensemble of Classification Trees
加载ionospheredata set.
loadionosphere
Train a bagged ensemble of 100 classification trees using all measurements.
rng(1)％可再现性mdl = fitCensemble（x，y，'Method','包')
Mdl = classreg.learning.classif.ClassificationBaggedEnsemble ResponseName: 'Y' CategoricalPredictors: [] ClassNames: {'b' 'g'} ScoreTransform: 'none' NumObservations: 351 NumTrained: 100 Method: 'Bag' LearnerNames: {'Tree'}理性终止：“在完成所需数量的训练周期后，通常终止。”fitinfo：[] fitinfodescription：'none'freSample：1替换：1 UseObsForlearner：[351x100逻辑]属性，方法，方法
MDL是aClassificationBaggedEnsemblemodel object.
MDL。训练是存储经过训练的分类树的100 x细胞向量的属性（compactClassificationTreemodel objects) that compose the ensemble.
绘制第一个受过训练的分类树的图。
查看（mdl.Trained {1}，'Mode','graph')
默认，fitcensemblegrows deep decision trees for bagged ensembles.
估计样本中的错误分类率。
l =重新公开（MDL）
L = 0
L是0，表明MDL非常适合分类培训数据。
提示
对于分类树的包装合奏，训练有素property ofens存储一个细胞向量en.compactClassificationTreemodel objects. For a textual or graphical display of treet在单元向量中，输入
查看（ens.trained {t}）
Extended Capabilities
在R2011a中引入
