[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fqRxcxG7IZT-wHipoqDFouVc1sZndb0SB76dEpsHk8IM":3},{"answer":4,"createTime":5,"id":6,"options":7,"origin":12,"question":19,"related":20,"source":30,"type":31},[],"2024-07-03 10:25:43",157203668,[8,9,10,11],"连续性属性可以采用最小最大值归一化,减少数值过大和不同属性量纲不同对网络的影响","对于非数值型的属性,可以使用独热(onehot)编码或模糊化转换成数值","BP神经网络的输入属性不需要筛选,因为网络本身有特征获取能力","预处理与否对神经网络的训练速度和分类准确率影响比较大",{"count":13,"courseId":14,"courseImg":15,"courseName":16,"workId":17,"workName":18},25,"d5146fdec329b210fdf75cd21ba93e22","https:\u002F\u002Ftihai-oss-cloud.itihey.com\u002Fimg\u002F48eba3395437dd881b1b9275a0ad1028.png","人工智能原理及应用","exam_122365533","第六章测试题","有关数据预处理对神经网络的影响,以下哪个说法是错误的",[21,32,41,50,59,68,77,80,90,99],{"answer":22,"createTime":5,"id":23,"options":24,"question":29,"source":30,"type":31},[],157203657,[25,26,27,28],"开销小、计算快","梯度为0时神经元难以正常工作","提高了网络训练速度","缓解了梯度消失问题","AlexNet使用ReLU激活函数的好处不包括以下哪个方面","v1",0,{"answer":33,"createTime":5,"id":34,"options":35,"question":40,"source":30,"type":31},[],157203660,[36,37,38,39],"减少深层网络的梯度消失问题","特征的重用","模型参数明显增加","增强特征的获取能力","ResNet的特点不包括以下哪一点",{"answer":42,"createTime":5,"id":43,"options":44,"question":49,"source":30,"type":31},[],157203663,[45,46,47,48],"学习率过大更容易导致训练陷入局部极小值","学习率可以随着训练误差动态调整效果更好","网络训练时刚开始学习率可以大一些,以便提高学习速度,随后应减少学习率,以免引起学习震荡","学习率可以与其他网络参数一起训练,对降低代价函数是有利的","有关神经网络训练时使用的学习率参数说法错误的是",{"answer":51,"createTime":5,"id":52,"options":53,"question":58,"source":30,"type":31},[],157203664,[54,55,56,57],"L2正则化","增加学习率","dropout","提前终止","下面哪个选项不是神经网络训练过程中过拟合的防止方法",{"answer":60,"createTime":5,"id":61,"options":62,"question":67,"source":30,"type":31},[],157203665,[63,64,65,66],"训练次数多,学习效率低,收敛速度慢","训练时新样本的加入对已经学习的结果没什么影响","易陷入局部极小值","隐层节点的选取缺乏理论指导","有关BP神经网络的说法,错误的是哪个",{"answer":69,"createTime":5,"id":70,"options":71,"question":76,"source":30,"type":31},[],157203666,[72,73,74,75],"1+Inx","1-y2","1-lnx","y(1-y)","sigmoid激活函数y=1\u002F(1+e-x)的导数是以下哪一个",{"answer":78,"createTime":5,"id":6,"options":79,"question":19,"source":30,"type":31},[],[8,9,10,11],{"answer":81,"createTime":82,"id":83,"options":84,"question":89,"source":30,"type":31},[],"2024-07-03 10:25:44",157203671,[85,86,87,88],"一个神经元可以有多个输入和多个输出","一个神经元可以拥有多个输入但只有一个输出","一个神经元只能拥有一个输入但可以有多个输出","一个神经元只能拥有一个输入和一个输出","下面关于单个神经元输入输出正确的是哪个",{"answer":91,"createTime":82,"id":92,"options":93,"question":98,"source":30,"type":31},[],157203673,[94,95,96,97],"KL散度","对数损失","交叉熵损失","均方误差","在BP神经网络中,用于分类的损失函数通常使用什么来衡量预测值和真实值之间的差异",{"answer":100,"createTime":82,"id":101,"options":102,"question":107,"source":30,"type":31},[],157203675,[103,104,105,106],"梯度消失","学习率选择不当","容易陷入局部最优解","输入数据维度过高","下列哪一项不是BP神经网络结构存在的问题"]