Skip to content

Commit

Permalink
修改错别字
Browse files Browse the repository at this point in the history
  • Loading branch information
richman authored and richman committed Jul 18, 2022
1 parent b58d779 commit efcc5d5
Show file tree
Hide file tree
Showing 3 changed files with 3 additions and 3 deletions.
2 changes: 1 addition & 1 deletion AI算法/machine-learning/RandomForest.md
Original file line number Diff line number Diff line change
Expand Up @@ -34,7 +34,7 @@

## 4. 为什么不用全样本训练?

全样本训练忽视了局部样本的规律(各个决策树趋于相同),对于模型的泛化能力是有害的,使随机森林算法在羊本层面失去了随机性
全样本训练忽视了局部样本的规律(各个决策树趋于相同),对于模型的泛化能力是有害的,使随机森林算法在样本层面失去了随机性

## 5. 为什么要随机特征?

Expand Down
2 changes: 1 addition & 1 deletion AI算法/machine-learning/metrics.md
Original file line number Diff line number Diff line change
Expand Up @@ -74,7 +74,7 @@ Acc与Error平等对待每个类别,即每一个样本判对 (0) 和判错 (1)

![image-20210616223053609](img/Metrics/image-20210616223053609.png)

### 准确率(查准率) Precision
### 精确率(查准率) Precision

Precision 是分类器预测的正样本中预测正确的比例,取值范围为[0,1],取值越大,模型预测能力越好。
$$
Expand Down
2 changes: 1 addition & 1 deletion AI算法/machine-learning/过拟合与欠拟合.md
Original file line number Diff line number Diff line change
Expand Up @@ -100,7 +100,7 @@

参考链接: https://www.zhihu.com/question/37096933/answer/475278057

## 7. Dropout为什么有助于过拟合?(重点)
## 7. Dropout为什么有助于防止过拟合?(重点)

* 取平均的作用

Expand Down

0 comments on commit efcc5d5

Please sign in to comment.