<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[github_39605284的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/github_39605284</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; github_39605284]]></copyright><item><title><![CDATA[leetcode-链表操作]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/121968910</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/121968910</guid><author>github_39605284</author><pubDate>Thu, 16 Dec 2021 10:27:02 +0800</pubDate><description><![CDATA[反转
部分反转
删除重复元素
删除倒数第k个节点
环形链表
环形链表入口
相交链表


]]></description><category></category></item><item><title><![CDATA[leetcode-第k大的数、最小的k个数]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/121955941</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/121955941</guid><author>github_39605284</author><pubDate>Wed, 15 Dec 2021 16:39:33 +0800</pubDate><description><![CDATA[剑指 Offer 40. 最小的k个数
class Solution:
    def getLeastNumbers(self, arr: List[int], k: int) -&gt; List[int]:
        def partition(arr, l, r):
            #选定中值
            pivotvalue = arr[l]
            lmark = l + 1
            rmark = r
            done ]]></description><category></category></item><item><title><![CDATA[leetcode--矩阵遍历与搜索]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/121949697</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/121949697</guid><author>github_39605284</author><pubDate>Wed, 15 Dec 2021 12:42:27 +0800</pubDate><description><![CDATA[54. 螺旋矩阵
class Solution(object):
    def spiralOrder(self, matrix):
        res = []
        row, col= len(matrix)-1, len(matrix[0])-1
        i, j = 0, 0

        while i&lt;=row and j &lt;=col:
            if i == row:
                for x in range(j, c]]></description><category></category></item><item><title><![CDATA[leetcode-二叉搜索树先关题目]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/121940450</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/121940450</guid><author>github_39605284</author><pubDate>Tue, 14 Dec 2021 21:24:38 +0800</pubDate><description><![CDATA[653. 两数之和 IV - 输入 BST
class Solution(object):
    def findTarget(self, root, k):
        def inorder(node):
            if not node:
                return []
            return inorder(node.left) + [node.val] + inorder(node.right)
        
        alist =]]></description><category></category></item><item><title><![CDATA[leetcode-零钱问题]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/121923241</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/121923241</guid><author>github_39605284</author><pubDate>Tue, 14 Dec 2021 11:37:37 +0800</pubDate><description><![CDATA[无限放取+方法数
arr里都是正数，没有重复，每一个值代表一种货币，每一种都可以用无限张，最终要找零钱是aim，
找零方法数放回
递归 入参 process(arr, index, rest)
自由使用arr[index…] 所有的面值
需要搞定的钱数rest
返回找零的方法数
终止条件，决定dp表的一些初始位置，非常重要
if index == arr.length:
    resturn rest == 1 if rest==0 else 0

递归方程
way = 0
zhang = 0
while]]></description><category></category></item><item><title><![CDATA[leetcode-同事总结的热门代码题]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/121919583</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/121919583</guid><author>github_39605284</author><pubDate>Tue, 14 Dec 2021 09:24:05 +0800</pubDate><description><![CDATA[最大公共子序列
最大公共子串
最长递增子序列
最长递增子串
连续子数组的最大和
连续子数组的最大乘积
最长回文子串
子数组和为k
连续子数组和为k

]]></description><category></category></item><item><title><![CDATA[leetcode-旋转数组相关题目解答]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/121908793</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/121908793</guid><author>github_39605284</author><pubDate>Mon, 13 Dec 2021 17:44:15 +0800</pubDate><description><![CDATA[33. 搜索旋转排序数组

153. 寻找旋转排序数组中的最小值

154. 寻找旋转排序数组中的最小值 II

[]

]]></description><category></category></item><item><title><![CDATA[leetcode之两、三、四、最近接之和题目]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/121903184</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/121903184</guid><author>github_39605284</author><pubDate>Mon, 13 Dec 2021 14:57:06 +0800</pubDate><description><![CDATA[1. 两数之和
class Solution(object):
    def twoSum(self, nums, target):
        d = {}
        for i in range(len(nums)):
            diff = target - nums[i]
            if diff in d:
                return [i, d[diff]]
            d[nums[i]] = i
        
    ]]></description><category></category></item><item><title><![CDATA[leetcode-合并有序数组、链表系列题目解法总结]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/121892429</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/121892429</guid><author>github_39605284</author><pubDate>Sun, 12 Dec 2021 19:55:09 +0800</pubDate><description><![CDATA[展示内容：一次遍历 将结果输出、而不是判断大小、再做其他冗余处理
88题 合并两个有序数组
版本一
class Solution:
    def merge(self , A, m, B, n):
        if n ==0: return A
        if m ==0: 
            A[:] = B[:]
            return A
        total, m, n = m+n-1, m-1, n-1
        while m&gt;=0 or n&]]></description><category></category></item><item><title><![CDATA[tf estimator debug]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/120884844</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/120884844</guid><author>github_39605284</author><pubDate>Thu, 21 Oct 2021 13:39:35 +0800</pubDate><description><![CDATA[How to print tensor in Tensorflow custom estimator for debugging?

]]></description><category></category></item><item><title><![CDATA[tf.newaxis vs tf.expand_dims]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/120306627</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/120306627</guid><author>github_39605284</author><pubDate>Wed, 15 Sep 2021 13:20:29 +0800</pubDate><description><![CDATA[There is no real difference between the three, but sometimes one or the other may be more convenient:
What’s the difference between using tf.expand_dims and tf.newaxis in Tensorflow?

]]></description><category></category></item><item><title><![CDATA[afm_layer]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/120236139</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/120236139</guid><author>github_39605284</author><pubDate>Sat, 11 Sep 2021 12:39:39 +0800</pubDate><description><![CDATA[# coding=utf-8
import itertools
import tensorflow as tf
from tensorflow.keras import Model
from tensorflow.keras.regularizers import l2
from tensorflow.keras.layers import Embedding, Dense, Dropout, Input

from tensorflow.keras.layers import Layer

class .]]></description><category></category></item><item><title><![CDATA[CIN_layer]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/120234766</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/120234766</guid><author>github_39605284</author><pubDate>Sat, 11 Sep 2021 11:03:45 +0800</pubDate><description><![CDATA[#coding=utf-8

import tensorflow as tf
from tensorflow.python.keras.layers import Concatenate,Conv1D,Reshape

## 只计算其中一层交叉的结果
def compressed_interaction_net(x0, xl, D, n_filters):
    """
    @param x0: 原始输入
    @param xl: 第l层的输入
    @param D: embedding d.]]></description><category></category></item><item><title><![CDATA[deep_cross_layer]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/120227980</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/120227980</guid><author>github_39605284</author><pubDate>Fri, 10 Sep 2021 18:49:49 +0800</pubDate><description><![CDATA[#coding=utf-8
import tensorflow as tf
from tensorflow.keras.layers import Dense, ReLU, Layer

class DeepCrossLater(Layer):
    def __init__(self,dim_stack, name=None):
        """
        :param hidden_unit: A list. Neural network hidden units.
        :pa]]></description><category></category></item><item><title><![CDATA[tf-矩阵相关知识]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/120140185</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/120140185</guid><author>github_39605284</author><pubDate>Mon, 06 Sep 2021 17:23:28 +0800</pubDate><description><![CDATA[1、向量 or 矩阵
    # 向量(3,)
    v1 = tf.constant([ 1, 2, 3])
    print(v1.shape)

    ## 矩阵(3, 1)
    v2 = tf.constant([ [1], [2], [3]])
    print(v2.shape)

2、最后一维升维
    ## 矩阵最后一维增加维度
    q = tf.constant([
                        [ 1, 2, 3],
                 ]]></description><category></category></item><item><title><![CDATA[机器学习概念理解（持续更新）]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/120079442</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/120079442</guid><author>github_39605284</author><pubDate>Fri, 03 Sep 2021 12:54:07 +0800</pubDate><description><![CDATA[Q:模型优化的直接目标为什么是loss
A：loss 是由损失函数决定的，损失函数定义了真实值与预测值之间的差距 loss代表了模型对数据拟合的程度，loss越小，说明模型对数据拟合得越好

]]></description><category></category></item><item><title><![CDATA[nc 跨机房传输文件]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/120013796</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/120013796</guid><author>github_39605284</author><pubDate>Tue, 31 Aug 2021 11:23:15 +0800</pubDate><description><![CDATA[本地机器：  nc -l 7001 &gt;  filename.zip
远端机器：nc  192.168.17.223  10013 &lt; filename.zip

]]></description><category></category></item><item><title><![CDATA[tf keras 训练过程简述]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/119949285</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/119949285</guid><author>github_39605284</author><pubDate>Fri, 27 Aug 2021 11:31:29 +0800</pubDate><description><![CDATA[持续更新。。。。。。。

]]></description><category></category></item><item><title><![CDATA[tf.keras.layer介绍]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/119908855</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/119908855</guid><author>github_39605284</author><pubDate>Wed, 25 Aug 2021 13:52:59 +0800</pubDate><description><![CDATA[通过继承tf.keras.layer
实现自定义层的最佳方法是继承tf.keras.Layer类并实现：

init ，在其中进行所有与输入无关的变量或常量的初始化
build，在其中知道输入张量的形状，并可以进行其余的初始化
call，在这里进行前向传播逻辑的计算的定义
请注意，不必等到build被调用才创建变量，也可以在__init__中直接定义。但是，在build中创建变量的好处在于，可以根据build的参数 input shape来自动调整变量的shape。而在__init__中创建变量意味着需要显]]></description><category></category></item><item><title><![CDATA[tf2.x函数签名]]></title><link>https://blog.youkuaiyun.com/github_39605284/article/details/119907366</link><guid>https://blog.youkuaiyun.com/github_39605284/article/details/119907366</guid><author>github_39605284</author><pubDate>Wed, 25 Aug 2021 11:28:15 +0800</pubDate><description><![CDATA[一般说法：
input_signature的好处：
1.可以限定函数的输入类型，以防止调用函数时调错，
2．一个函数有了input_signature之后，在tensorflow里边才可以保存成savedmodel。在保存成savedmodel的过程中，需要使用get_concrete_function函数把一个tf.function标注的普通的python函数变成带有图定义的函数。
使用经验：
当一个函数被@tf.function 比较后，可以使用get_concrete_function 然后传入原函数]]></description><category></category></item></channel></rss>