我有一个训练好的scikit-learn模型，它使用了多输出决策树（作为RandomForestRegressor）。没有对随机森林回归模型进行任何自定义配置来启用多输出行为，因为多输出行为是内置的。基本上，只要你将多输出训练数据拟合到模型中，模型就会在幕后切换到多输出模式。

此外，RandomForestRegressor是CoreML转换脚本支持的转换器。然而，在转换过程中，我遇到了以下错误和堆栈跟踪：

ValueError: 预期scikit-learn树中只有1个输出。

Traceback (most recent call last):  File "/Users/user0/Desktop/model_convert.py", line 7, in <module>    coreml_model = sklearn_to_ml.convert(model)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_converter.py", line 146, in convert    sk_obj, input_features, output_feature_names, class_labels = None)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_converter_internal.py", line 297, in _convert_sklearn_model    last_spec = last_sk_m.convert(last_sk_obj, current_input_features, output_features)._spec  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_random_forest_regressor.py", line 53, in convert    return _MLModel(_convert_tree_ensemble(model, feature_names, target))  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 195, in convert_tree_ensemble    scaling = scaling, mode = mode, n_classes = n_classes, tree_index = tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 68, in _recurse    _recurse(coreml_tree, scikit_tree, tree_id, left_child_id, scaling, mode, n_classes, tree_index)  File "/Library/Python/2.7/site-packages/coremltools/converters/sklearn/_tree_ensemble.py", line 75, in _recurse    raise ValueError('Expected only 1 output in the scikit-learn tree.')ValueError: Expected only 1 output in the scikit-learn tree.

以下是简单的转换代码：

from coremltools.converters import sklearn as sklearn_to_mlfrom sklearn.externals import joblibmodel = joblib.load("ms5000.pkl")print("正在转换模型")coreml_model = sklearn_to_ml.convert(model)print("正在保存CoreML模型")coreml_model.save("ms5000.mlmodel")

我该如何使CoreML转换脚本能够处理多输出决策树？是否可以在不完全重新编写新脚本的情况下对现有脚本进行修改？

回答：

CoreML目前还是一个全新的事物，目前还没有已知的第三方转换脚本来源。

coremltools文档的“模型”部分提供了如何使用Python生成CoreML模型的详细文档。也就是说，你可以使用文档中提供的模型接口将任何机器学习模型转换为CoreML模型。

目前，coremltools不支持多输出回归模型。如果你不想重新发明轮子，你需要通过引入一个新的输入来将模型转换为单输出模型，这个新输入对应于当前预测的输出是哪一个。

无论哪种方式，文档都在那里，应该可以帮助你开始。