【摘要】 切实提升学生评教的有效性和可信度是评教工作的关键点。对D高校数据从打分一致性、缺省修改率和评分奇异值三方面进行有效性探析，提出针对原始数据的降噪算法和基于K-means的分类评价方案。结果表明，降噪算法能够有效降低噪音数据对统计结果的影响，提高结果的可参考性。基于K-means的分类评价方案相较总分排名既能够充分发挥学生评教的积极作用，也能够有效降低非质量因素对结果的波动性影响。更多还原

【Abstract】 Improving the effectiveness and credibility of student evaluation results is key to carrying out student evaluation work. Based on real student evaluation data from D universities in recent years, a multidimensional data validity analysis was performed starting from scoring consistency, default value modification rate, and scoring strange value. We proposed a novel evaluation scheme based on K-means classification. Numerical results demonstrate that the proposed algorithm can effectively mitigate the influence of noisy data on analytical outcomes. Furthermore, the proposed evaluation scheme can not only give full play to the positive role of students′ evaluation, but also effectively reduce the effects of non-quality factors on the volatility of evaluation results.更多还原