[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fBwcicutrVgClaMOKjWfLuwKQzhsL4TeP_MiwAhASwjk":3},{"answer":4,"createTime":5,"id":6,"options":7,"origin":11,"question":15,"related":16,"source":25,"type":26},[],"2023-12-11 23:30:25",113184639,[8,9,10],"越复杂越好","越简单越好","适当限制其复杂程度尽可能利用所有特征",{"courseId":12,"courseImg":13,"courseName":14},"032297077570e209d28555f0a25f46e3","https:\u002F\u002Ftihai-oss-cloud.itihey.com\u002Fimg\u002F25b30343053994e8940089572d36015b.jpg","机器学习基础","决策树模型的规模应当是()",[17,27,30,39,48,57,66,75,84,93],{"answer":18,"createTime":5,"id":19,"options":20,"question":24,"source":25,"type":26},[],113184637,[21,22,23],"凹函数","凸函数","不确定","在一元线性回归中,均方差损失函数是. (单选题)","v1",0,{"answer":28,"createTime":5,"id":6,"options":29,"question":15,"source":25,"type":26},[],[8,9,10],{"answer":31,"createTime":5,"id":32,"options":33,"question":38,"source":25,"type":26},[],113184640,[34,35,36,37],"都相同","黄色","粉色","紫色","下图显示了三个逻辑回归模型的AUC-ROC曲线.不同的颜色表示不同超参数值的曲线.以下哪个AUC-ROC会给出最佳结果",{"answer":40,"createTime":5,"id":41,"options":42,"question":47,"source":25,"type":26},[],113184641,[43,44,45,46],"对数几率回归可以用来估计回归系数","对数几率回归可以用来度量模型拟合程度","对数几率回归是设计用来预测事件可能性的","以上所有","对数几率回归(logistics regression)和一般回归分析有什么区别",{"answer":49,"createTime":5,"id":50,"options":51,"question":56,"source":25,"type":26},[],113184642,[52,53,54,55],"2","1和2","1","都不是","下列哪种方法被用于预测因变量?( ) 1.线性回归 2.逻辑回归",{"answer":58,"createTime":5,"id":59,"options":60,"question":65,"source":25,"type":26},[],113184643,[61,62,63,64],"Word2Vec结果符合当前预料环境","Word2Vec受限于训练语料的数量和质量","Word2Vec基于概率统计","Word2Vec得到的都是语义上的同义词","中文同义词替换时,常用到Word2Vec,以下说法错误的是",{"answer":67,"createTime":5,"id":68,"options":69,"question":74,"source":25,"type":26},[],113184644,[70,71,72,73],"回归分析","主成份分析","聚类分析","分类分析","当不知道数据所带标签时,可以使用哪种技术促使带同类标签的数据与带其他标签的数据相分离",{"answer":76,"createTime":5,"id":77,"options":78,"question":83,"source":25,"type":26},[],113184646,[79,80,81,82],"正则化","归一化","交叉验证","失活层","若直接用初级学习器的训练集来产生次级训练集,则过拟合风险会比较大;一般会通过()等方式,用训练初级学习器未使用的样本来产生次级学习器的训练样本",{"answer":85,"createTime":5,"id":86,"options":87,"question":92,"source":25,"type":26},[],113184648,[88,89,90,91],"隐马模型(HMM)","朴素贝叶斯","支持向量机","LDA","以下哪种方法( )属于判别式模型(discriminative model)",{"answer":94,"createTime":5,"id":95,"options":96,"question":101,"source":25,"type":26},[],113184650,[97,98,99,100],"对数据取对数变化","对数据作归一化","都不对","对数据作主成分分析(PCA)和归一化","假设在训练中我们突然遇到了一个问题,在几次循环之后,误差瞬间降低.你认为数据有问题,于是你画出了数据并且发现也许是数据的偏度过大造成了这个问题,你打算怎么做来处理这个问题"]