The result problem

[prj-jxust]

亲爱的作者： 非常感谢您贡献的代码，并开放标注文件。但是我在评价中遇到了一些问题，首先按照提供的csv评价出来average acc的值为0.88作用，与论文提供的值不符，能否提供论文csv的值，还有就是论文中F-measure值根据公式2precisionRecall/(precision+Recall)，并不符合。希望作者指出一下我的疑惑，因为多分类的评价方式有好几种，请问您采用的是micro、Macro还是Weighted？真诚期待你的答复！

[YH2010]

Accuracy: 0.8800 Precision: 0.8651 Sensitivity: 0.8574 F1: 0.8606，This is the result based the csv author provided in the project.

[KeramatGH]

Accuracy: 0.8800 Precision: 0.8651 Sensitivity: 0.8574 F1: 0.8606，This is the result based the csv author provided in the project.

The trained model file has some errors. would you please send me the file (for research purpose) best regards [email protected]