The workaround is to fit TPOTClassifier in standalone mode, and create a PMMLPipeline object off the TPOTClassifier.fitted_pipeline_ ...
確定! 回上一頁