feedforward matlab,feedforwardnet与newff 归一化 神经网络

本文介绍了Matlab中创建和训练前馈神经网络feedforwardnet的用法,包括设置隐藏层大小和训练函数。默认情况下,输入和输出层的大小设置为0,可以通过`configure`进行手动配置。示例代码展示了如何使用feedforwardnet解决简单拟合问题的过程。
摘要由CSDN通过智能技术生成

function out1 = feedforwardnet(varargin)

%FEEDFORWARDNET Feedforward neural network.

%

%  Two (or more) layer feedforward networks can implement any finite

%  input-output function arbitrarily well given enough hidden neurons.

%

%  feedforwardnet(hiddenSizes,trainFcn) takes a 1xN vector of N hidden

%  layer sizes, and a backpropagation training function, and returns

%  a feed-forward neural network with N+1 layers.

%

%  Input, output and output layers sizes are set to 0.  These sizes will

%  automatically be configured to match particular data by train. Or the

%  user can manually configure inputs and outputs with configure.

%

%  Defaults are used if feedforwardnet is called with fewer arguments.

%  The default arguments are (10,'trainlm').

%

%  Here a feed-forward network is used to s

评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值