[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fE0h3d5IcQ938HQf6f4nzmQbaKtXyeMiVH6InZ_OiR3E":3},{"answer":4,"createTime":5,"id":6,"options":7,"origin":12,"question":19,"related":20,"source":30,"type":31},[],"2026-05-02 11:44:10",353681047,[8,9,10,11],"可以增加模型的复杂度","能够减少过拟合的风险","提高模型对噪声的鲁棒性","有助于模型快速收敛",{"count":13,"courseId":14,"courseImg":15,"courseName":16,"workId":17,"workName":18},83,"53e1d2ef4961cca8eea3e23969ad2cb9","https:\u002F\u002Ftihai-oss-cloud.itihey.com\u002Fimg\u002F03a579384a6dc297c89809b582fcc767.png","默认课程","work_52928797","第5章 经典卷积神经网络(上)(作业)","VGGNet中使用的ReLU激活函数有哪些优点",[21,32,41,49,58,67,76,85,94,103],{"answer":22,"createTime":5,"id":23,"options":24,"question":29,"source":30,"type":31},[],353681038,[25,26,27,28],"2014年","2012年","2015年","2013年","VGGNet是在哪一年由牛津大学计算机视觉组和Google DeepMind公司研究员共同研发出来的","v1",0,{"answer":33,"createTime":5,"id":34,"options":35,"question":40,"source":30,"type":31},[],353681039,[36,37,38,39],"降低特征图的维度","提高模型的泛化能力","减少模型参数","增加模型的非线性","VGGNet中的池化层主要用于什么目的",{"answer":42,"createTime":5,"id":43,"options":44,"question":48,"source":30,"type":31},[],353681040,[37,45,46,47],"捕获更多图像中的细节信息","简化模型结构","减少数据预处理时间","哪个选项不是数据增强可能带来的好处",{"answer":50,"createTime":5,"id":51,"options":52,"question":57,"source":30,"type":31},[],353681041,[53,54,55,56],"使用LRN","Dropout","增加网络深度","增大卷积核大小","AlexNet使用哪种方式来避免过拟合",{"answer":59,"createTime":5,"id":60,"options":61,"question":66,"source":30,"type":31},[],353681042,[62,63,64,65],"11×11, 5×5, 3×3","11×11, 3×3, 5×5","3×3, 5×5, 11×11","5×5, 11×11, 3×3","AlexNet中卷积核大小递减的顺序是什么",{"answer":68,"createTime":5,"id":69,"options":70,"question":75,"source":30,"type":31},[],353681043,[71,72,73,74],"3次","1次","2次","4次","LeNet-5网络结构中,卷积层和池化层的组合出现了几次",{"answer":77,"createTime":5,"id":78,"options":79,"question":84,"source":30,"type":31},[],353681044,[80,81,82,83],"在深层网络中表现更差","更容易导致梯度消失","提高了网络的训练速率","计算更复杂","AlexNet中引入的ReLU激活函数相较于Sigmoid的优势是什么",{"answer":86,"createTime":5,"id":87,"options":88,"question":93,"source":30,"type":31},[],353681045,[89,90,91,92],"池化层","全连接层","卷积层","输出层","LeNet-5的哪一层负责最终的分类输出",{"answer":95,"createTime":5,"id":96,"options":97,"question":102,"source":30,"type":31},[],353681046,[98,99,100,101],"作者发现LRN对性能没有显著提升","因为LRN会增加计算复杂度","VGGNet的结构设计使得LRN无法应用","LRN与ReLU激活函数不兼容","VGGNet为何没有在卷积层之间使用局部响应归一化(LRN)",{"answer":104,"createTime":5,"id":6,"options":105,"question":19,"source":30,"type":31},[],[8,9,10,11]]