摘要
由一名新闻记者-机器人与机器学习的工作人员新闻编辑每日新闻-人工智能的新数据在一份新的报告中呈现。根据NewsRx编辑来自加拿大埃德蒙顿的新闻报道，这项研究称：“通过学习分析和教育数据挖掘等技术，机器学习(ML)已经成为教育决策的一个整体。然而，在没有仔细检查的情况下采用机器学习工具可能会使偏见永久化。”新闻记者从阿尔伯塔大学的研究中获得了一句话:“尽管正在努力解决公平问题，但它们在教育数据集中的应用仍然有限。为了解决文献中提到的空白，本研究在一个教育数据集中评估了四种偏倚缓解技术的有效性,旨在预测学生辍学率。总体研究问题是:重加权、重采样、重采样和重采样技术的有效性如何。以学生的生物学性别作为受保护的属性，对这些技术的有效性进行了评估，发现重加权技术是无效的。结果与Baselin E条件相同。均匀和优先重采样技术都显著降低了预测偏差，特别是在FPR指标中，但代价是准确性和F1分数下降。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Fresh data on artificial intelligence are presented in a new report. According to news originating from Edmonton, Cana da, by NewsRx editors, the research stated, "Machine learning (ML) has become in tegral in educational decision-making through technologies such as learning anal ytics and educational data mining. However, the adoption of machine learning-dri ven tools without scrutiny risks perpetuating biases." The news reporters obtained a quote from the research from University of Alberta : "Despite ongoing efforts to tackle fairness issues, their application to educa tional datasets remains limited. To address the mentioned gap in the literature, this research evaluates the effectiveness of four bias mitigation techniques in an educational dataset aiming at predicting students' dropout rate. The overarc hing research question is: "How effective are the techniques of reweighting, res ampling, and Reject Option-based Classification (ROC) pivoting in mitigating the predictive bias associated with high school dropout rates in the HSLS:09 datase t?' The effectiveness of these techniques was assessed based on performance metr ics including false positive rate (FPR), accuracy, and F1 score. The study focus ed on the biological sex of students as the protected attribute. The reweighting technique was found to be ineffective, showing results identical to the baselin e condition. Both uniform and preferential resampling techniques significantly r educed predictive bias, especially in the FPR metric but at the cost of reduced accuracy and F1 scores."
NETL
NSTL