[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fGy2Scph5p6pU_2IaiLMKpT4oPrf0lwBnL-IcdM0XpXA":3},{"answer":4,"createTime":5,"id":6,"options":7,"origin":10,"question":13,"related":14,"source":20,"type":21},[],"2025-05-03 08:15:43",1062184942,[8,9],"对","错",{"courseImg":11,"courseName":12},"https:\u002F\u002Ftihai-oss-cloud.itihey.com\u002Fimg\u002Fd51bc44a4ff46d5688314c55e83b443c.jpg","深度学习","计算图执行前的传参过程中,函数调用帧是Session和执行器进行交互的窗口.( )",[15,22,27,32,37,42,52,57,62,67],{"answer":16,"createTime":5,"id":17,"options":18,"question":19,"source":20,"type":21},[],1062184896,[8,9],"代码tf.nn.sparse_softmax_cross_entropy_with_logits(labels,logits)的作用是定义加了稀疏的softmax交叉熵.( )","v2",3,{"answer":23,"createTime":5,"id":24,"options":25,"question":26,"source":20,"type":21},[],1062184898,[8,9],"Faster RCNN中生成推荐区域的算法为ROI Pooling.( )",{"answer":28,"createTime":5,"id":29,"options":30,"question":31,"source":20,"type":21},[],1062184900,[8,9],"TensorFlow中做深度学习计算的核心模块是tf.nn.( )",{"answer":33,"createTime":5,"id":34,"options":35,"question":36,"source":20,"type":21},[],1062184902,[8,9],"TensorFlow支持4种加载数据的方式.( )",{"answer":38,"createTime":5,"id":39,"options":40,"question":41,"source":20,"type":21},[],1062184903,[8,9],"在DCGAN中,判别器中使用的激活函数为LeakyReLu.( )",{"answer":43,"createTime":5,"id":44,"options":45,"question":50,"source":20,"type":51},[],1062184916,[46,47,48,49],"参数范数惩罚","提前终止训练","集成","增加模型的复杂度","深层人工神经网络具有很强的表示能力,但经常出现过拟合.为了提高神经网络的泛化能力,下列不能提高泛化能力的方法( )",0,{"answer":53,"createTime":5,"id":54,"options":55,"question":56,"source":20,"type":21},[],1062184918,[8,9],"TensorFlow提供了本地 分布式执行和分布式 本地执行两种执行方式.( )",{"answer":58,"createTime":5,"id":59,"options":60,"question":61,"source":20,"type":21},[],1062184920,[8,9],"TensorFlow中不同设备和机器间的通信都由Send和Recv节点进行,而Send和Recv使用Rendezvous机制完成数据交互Rendezvous是一个基于生产 消费者—消费 生产者模型设计的抽象类.( )",{"answer":63,"createTime":5,"id":64,"options":65,"question":66,"source":20,"type":21},[],1062184928,[8,9],"人工神经网络发展大致经历过(3)个阶段 ( )",{"answer":68,"createTime":5,"id":6,"options":69,"question":13,"source":20,"type":21},[],[8,9]]