通过实际缺陷样本测试,组合分类器的识别精度达到92%以上,对单个分类器识别准确率的提高幅度大于10%。
Tests on real strip defects show that the recognition accuracy reaches above 92%, and improves the accuracy of single classifier up to 10%.
最后,这段论述没有解释这样一种可能性,即调查样本本身带有缺陷。
Finally, this argument does not account for the possibility that the survey samples themselves were flawed.
事实上这是一个非典型的样本,不管是在抽样策略和它的组成上都是如此,而且统计管理也不能弥补这一缺陷。
This is a truly unrepresentative sample, both in sampling strategy and its make up, and statistical controls cannot compensate for this.
应用推荐