moyin/fzu-product
1
0
Fork 0
Code Issues 1 Pull Requests Packages Projects Releases Wiki Activity

fix: http to https

Browse Source
This commit is contained in:
camera-2018
2023-04-25 20:11:02 +08:00
parent 78828f70da
commit aa2d06bc00
19 changed files with 105 additions and 105 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
4.人工智能/ch02/ch2.1/ch2.1.5/TDM.md
View File

@@ -9,7 +9,7 @@
**树结构**
<div align=center>
<img src="http://ryluo.oss-cn-chengdu.aliyuncs.com/图片image-20220420213149324.png" alt="image-20210308142624189" style="zoom: 15%;" />
<img src="https://ryluo.oss-cn-chengdu.aliyuncs.com/图片image-20220420213149324.png" alt="image-20210308142624189" style="zoom: 15%;" />
</div>
如上图，树中的每一个叶子节点对应一个商品item，非叶子结点表示的是item的集合**（这里的树不限于二叉树）**。这种层次化结构体现了粒度从粗到细的item架构。
@@ -17,7 +17,7 @@
**整体结构**
<div align=center>
<img src="http://ryluo.oss-cn-chengdu.aliyuncs.com/图片image-20220420200433442.png" alt="image-20210308142624189" style="zoom: 70%;" />
<img src="https://ryluo.oss-cn-chengdu.aliyuncs.com/图片image-20220420200433442.png" alt="image-20210308142624189" style="zoom: 70%;" />
</div>
# 算法详解
@@ -34,7 +34,7 @@
2. 对兴趣进行建模
<div align=center>
<img src="http://ryluo.oss-cn-chengdu.aliyuncs.com/图片image-20220420214040264.png" alt="image-20210308142624189" style="zoom: 50%;" />
<img src="https://ryluo.oss-cn-chengdu.aliyuncs.com/图片image-20220420214040264.png" alt="image-20210308142624189" style="zoom: 50%;" />
</div>
​如上图，用户对叶子层item6感兴趣，可以认为它的兴趣是1，同层别的候选节点的兴趣为0，顺着着绿色线路上去的节点都标记为1，路线上的同层别的候选节点都标记为0。这样的操作就可以根据1和0构建用于每一层的正负样本。
@@ -44,7 +44,7 @@
3. 训练过程
<div align=center>
<img src="http://ryluo.oss-cn-chengdu.aliyuncs.com/图片image-20220420220831318.png" alt="image-20210308142624189" style="zoom: 15%;" />
<img src="https://ryluo.oss-cn-chengdu.aliyuncs.com/图片image-20220420220831318.png" alt="image-20210308142624189" style="zoom: 15%;" />
</div>
整体联合训练的方式如下：