![](https://img-blog.csdnimg.cn/20201014180756928.png?x-oss-process=image/resize,m_fixed,h_64,w_64)
机器学习
俺 也一样
这个作者很懒,什么都没留下…
展开
-
functools 中的 partial的作用
看到有一句代码如下:from functools import partialmy_dense_layer = partial(tf.layers.dense, activation=tf.nn.elu, kernel_initializer=he_init, ...原创 2019-12-11 21:48:19 · 676 阅读 · 0 评论 -
InceptionV3 迁移训练
InceptionV3.py:import tensorflow as tfimport osimport flower_photo_dispose as fdfrom tensorflow.python.platform import gfileprint("hello wrold1")model_path = "inception_dec_2015/"model_file =...原创 2019-11-02 13:35:33 · 203 阅读 · 0 评论 -
VGGNet 实现
import tensorflow as tffrom datetime import datetimeimport mathimport timebatch_size = 12num_batches = 100def myprint(conv): # print(conv.op.name, ' ', conv.get_shape().as_list()) str1 = st...原创 2019-10-31 21:51:15 · 280 阅读 · 0 评论 -
AlexNet 实现
import tensorflow as tfimport mathimport timefrom datetime import datetimebatch_size = 32num_batches = 100def myprint(conv): print(conv.op.name, " ", conv.get_shape().as_list())def infe...原创 2019-10-30 16:46:12 · 253 阅读 · 0 评论 -
ReLu和Sigmod函数
sigmod在网络较深的时候可能产生梯度弥散的问题relu函数则不会原创 2019-10-30 15:09:15 · 374 阅读 · 0 评论 -
LeNet-5 实现
import tensorflow as tfimport numpy as npfrom tensorflow.examples.tutorials.mnist import input_datamnist = input_data.read_data_sets("MNIST_data", one_hot=True)batch_size = 100learning_rate =...原创 2019-10-30 10:50:22 · 184 阅读 · 0 评论 -
(机器学习实战)第四章
都是在python3下面的:def loadDataSet(): postingList=[['my', 'dog', 'has', 'flea', 'problems', 'help', 'please'], ['maybe', 'not', 'take', 'him', 'to', 'dog', 'park', 'stupid'], ...原创 2019-09-15 08:09:05 · 673 阅读 · 1 评论 -
(机器学习实战)第三章
from math import logimport operatorfrom math import logimport operator#计算给定数据的香农熵def calcShannonEnt(dataSet): numEntries = len(dataSet) #得到长度 (行数)# print("numEntries :")# prin...原创 2019-09-15 08:09:00 · 292 阅读 · 0 评论 -
(机器学习实战)第七章
Jupyter NotebookUntitled最后检查: 5 小时前(自动保存)Current Kernel LogoPython 3 FileEditViewInsertCellKernelWidgetsHelpfrom numpy import *def loadSimpData(): datMat = matrix([[ 1. , 2.1], ...原创 2019-09-21 15:35:56 · 2420 阅读 · 0 评论