什么是“顺序”?感知器的
为那些知道答案的人提供一些简单的标记。
我现在正在复习考试,过去的问题之一是:
感知器的阶是什么意思?
我在讲义中找不到任何关于这方面的信息,甚至google也显得茫然。
我的猜测是顺序是神经网络中的层数,但这似乎不太正确。
A few simple marks for those who know the answer.
I'm doing revision for exams at the moment and one of the past questions is:
What is meant by the order of a perceptron?
I can't find any information about this in my lecture notes, and even google seems at a loss.
My guess is that the order is the number of layers in a neural network, but this doesn't seem quite right.
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(2)
如果您想评估多层神经网络的顺序或基数,您应该仅考虑内层的数量因为输入和输出层不被认为属于神经网络拓扑的基数。
例如,具有 2 个内层的 NN 的阶数=2。
最有趣的是,大多数时候,超过一层对于性能和训练都没有用处。
If you want to evaluate the order or cardinality of a multilayered NN you should consider just the number of inner layer as input and output layer are not considered belonging to the cardinality of the NN topology.
For example a NN with 2 inner layer has order=2.
The funniest thing is that more than one layer is, most of the times, unusefull neither for performance neither for training.
学习算法中的近似阶。请参阅优化算法的顺序。
Order of approximation in the learning algorithm. See orders of optimization algorithms.