没有显示决策树结果

AgusKrisn4

当我运行它时，它只显示性能向量、属性权重和示例集，没有显示决策树结果。在我使用优化选择(进化)之前，它工作得很好

这是我的模型截图


这里是优化选择(进化)


这里是分割验证


在这里我的XML

<？xml version = " 1.0 " encoding = " utf - 8 " ?> <过程version = " 9.10.001”>

> <上下文

<输入/ >

<输出/ >

<宏/ >

> < /上下文

<参数键= " logverbosity " value = " init " / >

<参数键= " random_seed " value = " 2001 " / >

<参数键= " send_mail " value = "永远" / >

<参数键= " notification_email“价值= " / >

<参数键= " process_duration_for_mail " value = " 30 " / >

<参数键=“编码”值= "系统" / >

<过程扩展= " true " >

<参数键=“csv_file”值= " D: \ Kuliah \ SKRIPSI \ data.csv " / >

<参数键=“column_separators”值= ";" / >

<参数键= " trim_lines " value = " false " / >

<参数键= " use_quotes " value = " true " / >

<参数键= value =“quotes_character“;" / >

<参数键= " escape_character " value = " \ " / >

<参数键= " skip_comments " value = " true " / >

<参数键= " comment_characters " value = " # " / >

<参数键= " starting_row " value = " 1 " / >

<参数键= " parse_numbers " value = " true " / >

< = value =“decimal_character参数的关键。" / >

<参数键= " grouped_digits " value = " false " / >

<参数键= " grouping_character“价值= "、" / >

<参数键= " infinity_representation“价值= " / >

<参数键= " date_format“价值= " / >

<参数键= " first_row_as_names " value = " true " / >

<列出关键= "注释" / >

<参数键= " time_zone " value = "系统" / >

<参数键=“编码”值= " windows - 1252 " / >

<参数键= " read_all_values_as_polynominal " value = " false " / >

<列出关键= " data_set_meta_data_information " >

<参数键= " 1 " value = " K / L / PD.true.polynominal.attribute " / >

<参数键= " 2 " value = " HPS.true.integer.attribute " / >

<参数键= " 7 " value = " Status.true.polynominal.attribute " / >

< / >列表

<参数键= " read_not_matching_values_as_missings " value = " false " / >

< /操作符>

<参数键= " attribute_name " value = "地位" / >

<参数键= " target_role " value = "标签" / >

<列出关键= " set_additional_roles " / >

< /操作符>

<操作符激活="true" class="optimize_selection_evolutionary" compatibility="9.10.001" expanded="true" height="103" name="Optimize Selection (Evolutionary)" width="90" x="447" y="34">

<参数键= " use_exact_number_of_attributes " value = " false " / >

<参数键= " restrict_maximum " value = " false " / >

<参数键= " min_number_of_attributes " value = " 1 " / >

<参数键= " max_number_of_attributes " value = " 1 " / >

<参数键= " exact_number_of_attributes " value = " 1 " / >

<参数键= " initialize_with_input_weights " value = " false " / >

<参数键= " population_size " value = " 5 " / >

<参数键= " maximum_number_of_generations " value = " 30 " / >

<参数键= " use_early_stopping " value = " false " / >

<参数键= " generations_without_improval " value = " 2 " / >

<参数键= " normalize_weights " value = " true " / >

<参数键= " use_local_random_seed " value = " false " / >

<参数键= " local_random_seed " value = " 1992 " / >

<参数键= " user_result_individual_selection " value = " false " / >

<参数键= " show_population_plotter " value = " false " / >

<参数键= " plot_generations " value = " 10 " / >

<参数键= " constraint_draw_range " value = " false " / >

<参数键= " draw_dominated_points " value = " true " / >

<参数键= " maximal_fitness " value = "无限" / >

<参数键= " selection_scheme " value = "比赛" / >

<参数键= " tournament_size " value = " 0.25 " / >

<参数键= " start_temperature " value = " 1.0 " / >

<参数键= " dynamic_selection_pressure " value = " true " / >

<参数键= " keep_best_individual " value = " false " / >

<参数键= " save_intermediate_weights " value = " false " / >

<参数键= " intermediate_weights_generations " value = " 10 " / >

<参数键= " p_initialize " value = " 0.5 " / >

<参数键= " p_mutation " value = " -1.0 " / >

<参数键= " p_crossover " value = " 0.5 " / >

<参数键= " crossover_type " value = "统一" / >

<过程扩展= " true " >

<参数键= " create_complete_model " value = " false " / >

<参数键=“分裂”值= "相对" / >

<参数键= " split_ratio " value = " 0.7 " / >

<参数键= " training_set_size " value = " 100 " / >

<参数键= " test_set_size " value = " 1 " / >

<参数键= " sampling_type " value = "自动" / >

<参数键= " use_local_random_seed " value = " false " / >

<参数键= " local_random_seed " value = " 1992 " / >

<过程扩展= " true " >

<参数键= "标准" value = " gain_ratio " / >

<参数键= " maximal_depth " value = " 10 " / >

<参数键= " apply_pruning " value = " true " / >

<参数键= "信心" value = " 0.1 " / >

<参数键= " apply_prepruning " value = " true " / >

<参数键= " minimal_gain " value = " 0.01 " / >

<参数键= " minimal_leaf_size " value = " 2 " / >

<参数键= " minimal_size_for_split " value = " 4 " / >

<参数键= " number_of_prepruning_alternatives " value = " 3 " / >

< /操作符>

< portSpacing端口= " source_training”间隔= " 0 " / >

< portSpacing端口= " sink_model”间隔= " 0 " / >

> < /过程

<过程扩展= " true " >

<列出关键= " application_parameters " / >

<参数键= " create_view " value = " false " / >

< /操作符>

<参数键= " main_criterion " value = "第一次" / >

<参数键=“准确性”价值= " true " / >

<参数键= " classification_error " value = " false " / >

<参数键=“卡巴”值= " false " / >

<参数键= " weighted_mean_recall " value = " false " / >

<参数键= " weighted_mean_precision " value = " false " / >

<参数键= " spearman_rho " value = " false " / >

<参数键= " kendall_tau " value = " false " / >

<参数键= " absolute_error " value = " false " / >

<参数键= " relative_error " value = " false " / >

<参数键= " relative_error_lenient " value = " false " / >

<参数键= " relative_error_strict " value = " false " / >

<参数键= " normalized_absolute_error " value = " false " / >

<参数键= " root_mean_squared_error " value = " false " / >

<参数键= " root_relative_squared_error " value = " false " / >

<参数键= " squared_error " value = " false " / >

<参数键=“相关性”值= " false " / >

<参数键= " squared_correlation " value = " false " / >

<参数键=“叉”值= " false " / >

<参数键=“保证金”值= " false " / >

<参数键= " soft_margin_loss " value = " false " / >

<参数键= " logistic_loss " value = " false " / >

<参数键= " skip_undefined_labels " value = " false " / >

<参数键= " use_example_weights " value = " true " / >

<列出关键= " class_weights " / >

< /操作符>

< portSpacing端口= " source_model”间隔= " 0 " / >

> < /过程

< /操作符>

< portSpacing端口= " sink_performance”间隔= " 0 " / >

> < /过程

< /操作符>

> < /过程

< /操作符>

> < /过程

BalazsBarany

这就跟你问声好！

特征选择不应该显示模型。它尝试了很多不同的模式。

只需将Decision Tree放在选择之后，它将基于所选属性构建。

最好的方法是将特征选择放入交叉验证(是的，这样你就有了“外部”和“内部”验证)。通过这种方式，您将对整个建模过程的准确性得到一个很好的估计。

在交叉验证中:

左:特征选择(进化)，决策树
右:应用模型、性能

问候,
Balazs

Jerwuney

你好,

你所使用的操作符将为你提供最重要的特征，你想使用数据来做什么;分类、回归等。所以你必须在输出处使用新的数据集用决策树做你的预测。

或者您可以使用建模>优化>参数下的“优化参数(进化)”，它将另外给出树模型。我希望能有所帮助。

的问候。