[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$f4uJVP-TS7O8Q6g5VFORl3LU0bjPqMchQstQSESEpKzg":3},{"answer":4,"createTime":5,"id":6,"options":7,"origin":10,"question":14,"related":15,"source":26,"type":108},[],"2023-12-28 19:21:19",123371501,[8,9],"正确","错误",{"courseId":11,"courseImg":12,"courseName":13},"15706eb2a9c7ceda1f4403a1c0353f39","https:\u002F\u002Ftihai-oss-cloud.itihey.com\u002Fimg\u002F9311a73ad2b70ba131662b89e7ebd774.png","机器学习","任何情况下,我们说,神经网络学习算法都优于支持向量机学习算法",[16,28,38,48,58,67,76,85,94,99],{"answer":17,"createTime":18,"id":19,"options":20,"question":25,"source":26,"type":27},[],"2023-12-30 21:04:58",123370608,[21,22,23,24],"对数几率函数","对数线性函数","Sigmoid 函数","单位阶跃函数","对数几率回归的联系函数是( )","v1",0,{"answer":29,"createTime":30,"id":31,"options":32,"question":37,"source":26,"type":27},[],"2023-12-30 21:04:57",123370609,[33,34,35,36],"版本空间","假设空间","输出空间","输入空间","反映事件或对象在某方面的表现或性质的事项,称为"属性";属性上的取值,称为"属性值";属性张成的空间称为( )",{"answer":39,"createTime":40,"id":41,"options":42,"question":47,"source":26,"type":27},[],"2023-12-30 12:07:55",123370612,[43,44,45,46],"C4.5算法","标准BP算法","累积BP算法","ID3算法","( )每次更新只针对单个样例,参数更新的非常频繁,而且对不同样例进行更新的效果可能出现"抵消"现象",{"answer":49,"createTime":50,"id":51,"options":52,"question":57,"source":26,"type":27},[],"2023-12-28 19:20:03",123370613,[53,54,55,56],"以上都是","非","与","或","感知机由两层神经元组成,输入层接收外界输入信号后传递给输出层,输出层是M-P神经元,亦称"阈值逻辑单元".它能容易的实现逻辑( )运算",{"answer":59,"createTime":30,"id":60,"options":61,"question":66,"source":26,"type":27},[],123370615,[62,63,64,65],"连接主义学派","频率主义学派","符号主义学派","贝叶斯学派","( )认为参数虽然未知,但却是客观存在的固定值,因此,可通过优化似然函数等准则来确定参数值",{"answer":68,"createTime":18,"id":69,"options":70,"question":75,"source":26,"type":27},[],123370616,[71,72,73,74],"误差逆传播算法","CART决策树算法","C4.5决策树算法","ID3决策树算法","给定划分属性,若样本在该属性上的值缺失,则让该样本以不同的概率划入到不同的子结点中去. ( )使用了上述解决方案",{"answer":77,"createTime":50,"id":78,"options":79,"question":84,"source":26,"type":27},[],123370617,[80,81,82,83],"基尼指数","增益率","信息增益","信息熵","著名的C4.5决策树学习算法就是以( )为准则来选择划分属性",{"answer":86,"createTime":40,"id":87,"options":88,"question":93,"source":26,"type":27},[],123370618,[89,90,91,92],"点估计","面估计","概率估计","极大似然估计","源自频率主义学派的( ),是根据数据采样来估计概率分布参数的经典方法",{"answer":95,"createTime":18,"id":96,"options":97,"question":98,"source":26,"type":27},[],123370622,[34,33,35,36],"现实问题中我们常面临很大的假设空间,但学习过程是基于有限样本训练集进行的,因此,可能有多个假设与训练集一致,即存在着一个与训练集一致的"假设集合",我们称之为( )",{"answer":100,"createTime":40,"id":101,"options":102,"question":107,"source":26,"type":27},[],123370623,[103,104,105,106],"p\u002F(1-p)","p","(1-p)\u002Fp","1-p","假设某事件发生的概率为p,则此事件发生的几率为( )",3]