get_sample_class_mlpT_get_sample_class_mlpGetSampleClassMlpGetSampleClassMlpget_sample_class_mlp (算子)

名称

get_sample_class_mlpT_get_sample_class_mlpGetSampleClassMlpGetSampleClassMlpget_sample_class_mlp — 从多层感知机的训练数据中返回训练样本。

签名

get_sample_class_mlp( : : MLPHandle, IndexSample : Features, Target)

描述

get_sample_class_mlpget_sample_class_mlpGetSampleClassMlpGetSampleClassMlpGetSampleClassMlpget_sample_class_mlp reads out a training sample from the multilayer perceptron (MLP) given by MLPHandleMLPHandleMLPHandleMLPHandleMLPHandlemlphandle that was added with add_sample_class_mlpadd_sample_class_mlpAddSampleClassMlpAddSampleClassMlpAddSampleClassMlpadd_sample_class_mlp or read_samples_class_mlpread_samples_class_mlpReadSamplesClassMlpReadSamplesClassMlpReadSamplesClassMlpread_samples_class_mlp。The index of the sample is specified with IndexSampleIndexSampleIndexSampleIndexSampleindexSampleindex_sample. The index is counted from 0, i.e., IndexSampleIndexSampleIndexSampleIndexSampleindexSampleindex_sample must be a number between 0 and NumSamplesNumSamplesNumSamplesNumSamplesnumSamplesnum_samples - 1, where NumSamplesNumSamplesNumSamplesNumSamplesnumSamplesnum_samples can be determined with get_sample_num_class_mlpget_sample_num_class_mlpGetSampleNumClassMlpGetSampleNumClassMlpGetSampleNumClassMlpget_sample_num_class_mlp。The training sample is returned in FeaturesFeaturesFeaturesFeaturesfeaturesfeatures and TargetTargetTargetTargettargettarget. FeaturesFeaturesFeaturesFeaturesfeaturesfeatures is a feature vector of length NumInputNumInputNumInputNumInputnumInputnum_input, while TargetTargetTargetTargettargettarget is a target vector of length NumOutputNumOutputNumOutputNumOutputnumOutputnum_output (see add_sample_class_mlpadd_sample_class_mlpAddSampleClassMlpAddSampleClassMlpAddSampleClassMlpadd_sample_class_mlp and create_class_mlpcreate_class_mlpCreateClassMlpCreateClassMlpCreateClassMlpcreate_class_mlp).

get_sample_class_mlpget_sample_class_mlpGetSampleClassMlpGetSampleClassMlpGetSampleClassMlpget_sample_class_mlp can, for example, be used to reclassify the training data with classify_class_mlpclassify_class_mlpClassifyClassMlpClassifyClassMlpClassifyClassMlpclassify_class_mlp in order to determine which training samples, if any, are classified incorrectly.

执行信息

多线程类型：可重入（与非独占算子并行运行）。
多线程作用域：全局（可从任何线程调用）。
未采用并行化处理。

参数

MLPHandleMLPHandleMLPHandleMLPHandleMLPHandlemlphandle (输入控制) class_mlp → (handle)

MLP 句柄。

IndexSampleIndexSampleIndexSampleIndexSampleindexSampleindex_sample (输入控制) integer → (integer)

Number of stored training sample.

FeaturesFeaturesFeaturesFeaturesfeaturesfeatures (输出控制) real-array → (real)

训练样本的特征向量。

TargetTargetTargetTargettargettarget (输出控制) real-array → (real)

Target vector of the training sample.

示例（HDevelop）

* Train an MLP
create_class_mlp (NumIn, NumHidden, NumOut, 'softmax', \
                  'canonical_variates', NumComp, 42, MLPHandle)
read_samples_class_mlp (MLPHandle, 'samples.mtf')
train_class_mlp (MLPHandle, 100, 1, 0.01, Error, ErrorLog)
* Reclassify the training samples
get_sample_num_class_mlp (MLPHandle, NumSamples)
for I := 0 to NumSamples-1 by 1
    get_sample_class_mlp (MLPHandle, I, Data, Target)
    classify_class_mlp (MLPHandle, Data, 1, Class, Confidence)
    Result := gen_tuple_const(NumOut,0)
    Result[Class] := 1
    Diffs := Target-Result
    if (sum(fabs(Diffs)) > 0)
        * Sample has been classified incorrectly
    endif
endfor

结果

如果参数有效，算子 get_sample_class_mlpget_sample_class_mlpGetSampleClassMlpGetSampleClassMlpGetSampleClassMlpget_sample_class_mlp 返回值 2 ( H_MSG_TRUE )。如有必要，则抛出异常。

可能的前趋

add_sample_class_mlpadd_sample_class_mlpAddSampleClassMlpAddSampleClassMlpAddSampleClassMlpadd_sample_class_mlp, read_samples_class_mlpread_samples_class_mlpReadSamplesClassMlpReadSamplesClassMlpReadSamplesClassMlpread_samples_class_mlp, get_sample_num_class_mlpget_sample_num_class_mlpGetSampleNumClassMlpGetSampleNumClassMlpGetSampleNumClassMlpget_sample_num_class_mlp

可能的后继

classify_class_mlpclassify_class_mlpClassifyClassMlpClassifyClassMlpClassifyClassMlpclassify_class_mlp, evaluate_class_mlpevaluate_class_mlpEvaluateClassMlpEvaluateClassMlpEvaluateClassMlpevaluate_class_mlp

另见

create_class_mlpcreate_class_mlpCreateClassMlpCreateClassMlpCreateClassMlpcreate_class_mlp

模块

基础

算子