![](https://img-blog.csdnimg.cn/20201014180756927.png?x-oss-process=image/resize,m_fixed,h_64,w_64)
卷积神经网络
北漂客
这个作者很懒,什么都没留下…
展开
-
numpy实现卷积网络
import numpy as npclass conv(): def __init__(self,in_channles,out_channles,kernel=3,stride=1,padding=0): self.in_channles = in_channles self.out_channles =out_channles self.kernel = kernel self.stride = stride .原创 2021-08-05 11:03:04 · 337 阅读 · 0 评论 -
numpy实现BN层
BN层的主要逻辑是将数据进行一个标准化,然后添加了,两个可以学习的参数,来减弱标准化。class BN(): def __init__(self,channel,weight=[],bias=[],eps=1e-05): self.channel = channel self.weight = weight self.bias = bias self.eps =eps self.running_mean = [].原创 2021-08-05 10:58:55 · 618 阅读 · 0 评论 -
numpy实现最大池化
最大池化的逻辑是在一个固定大小的格子类找到最大的值,他是一个无权重层class max_pool(): def __init__(self,pooling=(2,2), strides=(2, 2), padding=(0, 0)): self.pooling = pooling self.strides = strides self.padding = padding def forward(self,z): z =原创 2021-08-05 10:56:34 · 947 阅读 · 0 评论 -
numpy实现cnn层
cnn的原理可以查看百度的一些介绍,主要是利用划窗进行矩阵的对应相乘运算class conv(): def __init__(self,in_channles,out_channles,kernel=3,stride=1,padding=0): self.in_channles = in_channles self.out_channles =out_channles self.kernel = kernel self.stri.原创 2021-08-05 10:30:23 · 303 阅读 · 0 评论 -
numpy实现全连接
全连接层可以看作是矩阵乘法。class Liner(): def __init__(self,in_features,out_features,bias=0): self.in_features =in_features self.out_features = out_features self.weight = np.zeros((in_features,out_features)) self.bias = np.zeros原创 2021-08-05 10:23:04 · 252 阅读 · 0 评论