[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fxuaZCH2EVhe0HiG1AQ5YIF2qnEnK82bOl-ULj6SMyoQ":3},{"answer":4,"createTime":5,"id":6,"options":7,"origin":12,"question":16,"related":17,"source":28,"type":29},[],"2023-12-10 14:10:33",112631371,[8,9,10,11],"2和4","1和3","2","1",{"courseId":13,"courseImg":14,"courseName":15},"032297077570e209d28555f0a25f46e3","https:\u002F\u002Ftihai-oss-cloud.itihey.com\u002Fimg\u002F25b30343053994e8940089572d36015b.jpg","机器学习基础","假设我们要解决一个二类分类问题, 我们已经建立好了模型, 输出是0或1, 初始时设阈值为0.5, 超过0.5概率估计, 就判别为1, 否则就判别为0 ; 如果我们现在用另一个大于0.5的阈值, 那么现在关于模型说法, 正确的是 1模型分类的召回率会降低或不变 2.模型分类的召回率会升高 3.模型分类准确率会升高或不变 4.模型分类准确率会降低",[18,30,39,48,57,65,74,82,91,100],{"answer":19,"createTime":20,"id":21,"options":22,"question":27,"source":28,"type":29},[],"2023-12-10 14:10:29",112631215,[23,24,25,26],"以上均不正确","误分类","不确定","正确分类","假定你使用SVM学习数据X,数据X里面有些点存在错误.现在如果你使用一个二次核函数,多项式阶数为2,使用松弛变量C作为超参之一.如果使用较小的C(C趋于0),则","v1",0,{"answer":31,"createTime":20,"id":32,"options":33,"question":38,"source":28,"type":29},[],112631216,[34,35,36,37],"朴素贝叶斯","决策树","SVM","逻辑回归","()理论的核心思想,就是选择最高概率的决策",{"answer":40,"createTime":20,"id":41,"options":42,"question":47,"source":28,"type":29},[],112631217,[43,44,45,46],"有放回地从总共N个样本中抽样n个样本","无放回地从总共N个样本中抽样n个样本","无放回地从总共M个特征中抽样m个特征","有放回地从总共M个特征中抽样m个特征","bootstrap数据是什么意思?(提示:考"bootstrap"和"boosting"区别) (单选题)",{"answer":49,"createTime":20,"id":50,"options":51,"question":56,"source":28,"type":29},[],112631219,[52,53,54,55],"在adaboost算法中,所有被分错样本的权重更新比例不相同","boosting和bagging都是组合多个分类器投票的方法,二者都是根据单个分类器的正确率确定其权重","SVM对噪声(如来自其他分部的噪声样本)具备鲁棒性","给定n个数据点,如果其中一半用于训练,一半用户测试,则训练误差和测试误差之间的差别会随着n的增加而减少的","以下说法中错误的是 (单选题)",{"answer":58,"createTime":20,"id":59,"options":60,"question":64,"source":28,"type":29},[],112631221,[61,62,63,35],"adboost","xgboost","随机森林","下列不属于集成学习算法()",{"answer":66,"createTime":20,"id":67,"options":68,"question":73,"source":28,"type":29},[],112631224,[69,70,71,72],"2x-y=0","x+2y=3","x+2y=5","2x+y=4","有两个样本点,第一个点为正样本,它的特征向量是(0,-1);第二个点为负样本,它的特征向量是(2,3),从这两个样本点组成的训练集构建一个线性SVM分类器的分类面方程是",{"answer":75,"createTime":20,"id":76,"options":77,"question":81,"source":28,"type":29},[],112631226,[37,78,79,80],"线性回归","以上说法都不对","线性回归和逻辑回归都行","一般来说,下列哪种方法常用来预测连续独立变量?( ) (单选题)",{"answer":83,"createTime":20,"id":84,"options":85,"question":90,"source":28,"type":29},[],112631228,[86,87,88,89],"学习—Baum-Welch算法","学习—前向后向算法","评估—前向后向算法","解码—维特比算法","隐马尔可夫模型三个基本问题以及相应的算法说法错误的是",{"answer":92,"createTime":20,"id":93,"options":94,"question":99,"source":28,"type":29},[],112631230,[95,96,97,98],"直方图","都不对","条形图","散点图","为了检验连续变量x,y之间的线性关系,下列哪种图最合适? (单选题)",{"answer":101,"createTime":20,"id":102,"options":103,"question":105,"source":28,"type":29},[],112631231,[104,35,37,78],"聚类","机器学习分为:监督学习,无监督学习,半监督学习.其中监督学习是通过训练让机器自己找到特征和标签之间的联系,在以后面对只有特征而没有标签的数据时可以自己判别出标签.下列哪项不属于监督学习的算法()"]