Skip to content

Commit

Permalink
fix bug
Browse files Browse the repository at this point in the history
  • Loading branch information
shataowei committed Dec 21, 2019
1 parent 52a616a commit d9f3046
Show file tree
Hide file tree
Showing 3 changed files with 177 additions and 43 deletions.
213 changes: 172 additions & 41 deletions .idea/workspace.xml

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

4 changes: 2 additions & 2 deletions 机器学习/集成学习/GBDT.md
Original file line number Diff line number Diff line change
Expand Up @@ -14,8 +14,8 @@
adaboost,gbdt等等

# gbdt和boostingtree的boosting分别体现在哪里?
- boostingtree利用基模型学习器,拟合的是当前模型与标签值的残差
- gbdt利用基模型学习器,拟合的是当前模型与标签值的残差的负梯度
- boostingtree利用基模型学习器,拟合的是mse(回归)或者指数损失函数(分类)
- gbdt利用基模型学习器,拟合的是当前模型与标签值的损失函数的负梯度

# gbdt的中的tree是什么tree?有什么特征?
Cart tree,但是都是回归树
Expand Down
3 changes: 3 additions & 0 deletions 自然语言处理/GloVe.md
Original file line number Diff line number Diff line change
Expand Up @@ -8,3 +8,6 @@
- GloVe算法本身使用了全局信息,自然内存费的也就多一些
- 公现矩阵,NXN的,N为词袋量
- W2V的工程实现结果相对来说支持的更多,比如most_similarty等功能

# 如何处理未出现词?
按照词性进行已知词替换,\[unknow-n],\[unknow-a],\[unknow-v]...,然后再进行训练。实际去用的时候,判断词性后直接使用对应的unknown-?向量替代

0 comments on commit d9f3046

Please sign in to comment.