基于贝叶斯的混合式教学效果溯源分析——以气象学与气候学为例  

Bayesian-based traceability analysis of mixed teaching effect:Take meteorology and climatology

作　　者：程乖梅[1] 张代青[1] 沈春颖[1] 左黔 何士华[1] CHENG Guaimei;ZHANG Daiqing;SHEN Chunying;ZUO Qian;HE Shihua(Faculty of Social Science,Kunming University of Science and Technology,Kunming Yunnan 650000,China)

机构地区：[1]昆明理工大学电力工程学院,云南昆明650000

出　　处：《未来与发展》2024年第11期68-73,共6页Future and Development

基　　金：昆明理工大学“线上线下混合式一流本科课程”建设项目;中国水利学会-基于工程教育专业认证背景下水利高等教育教学改革研究课题“后疫情时代下基于‘新工科+OBE理念’的西南地方高校水利类专业实践教学改革”。

摘　　要：在后疫情时代,混合式教学成为高校教学模式改革的热点。针对混合式教学效果评价中的不确定性问题,根据影响教学效果的因素之间的互信息,利用最大支撑树构建初始贝叶斯网络结构;采用广度优先搜索预先确定节点序,将节点序输入K2算法优化贝叶斯网络结构;应用GeNIe软件建立优化后的贝叶斯网络结构模型,进行贝叶斯先验推理和后验仿真推理。结果表明:该模型在先验条件下和实测数据拟合良好;在后验条件下,提高慕课学习和课程学习报告得分秀率可取得良好的教学效果。该模型清楚地表达了教学方式与教学效果之间复杂的内在联系,直观定量地反映其变化规律,可为教学改进提供指导。In the post-epidemic era,Mixed teaching has become a hot spot in the reform of teaching mode in colleges and universities.Aiming at the uncertainty in the evaluation of mixed teaching effect,The initial Bayesian network structure is constructed by using maximum spanning tree,According to the mutual information between the factors that affect the teaching effect,Breadth-first search is used to determine node order.The node order is input K2 algorithm to optimize the Bayesian network structure.Genie software is used to build an optimized Bayesian network model.A priori reasoning and a posteriori simulation show that the model fits the measured data well under the prior condition and good teaching results can be obtained by improving the rate of excellence in MOOC and the excellent rate of course report score under the posteriori condition.The complex internal relations between teaching methods and teaching effects is shown clearly by the model.The changing law is reflected directly and quantitatively.The guidance is provided to improve teaching.

关 键 词：混合式教学 教学效果溯源分析 贝叶斯网络结构 

分 类 号：G648.1[文化科学—高等教育学]