# 前言

(炼金何尝不是呢？各种配方温度时间等等的调整)

# 如何Debug

• 寻找更多的数据
• 让网络层数更深一些
• 在神经网络中采取一些新的方法
• 训练的时间更长点(更多的迭代次数)
• 改变batch-size
• 尝试使用正则化技术(权重衰减)
• 权衡结果的偏置和方差(bias and variance)
• 使用更多的GPU

## 超参数

• 学习速率(如何设置学习率)
• batchsize
• 权重衰减系数
• dropout系数
• 选择适用的优化器
• 是否使用batch-normalization
• 神经网络设计的结构(比如神经网络的层数，卷积的大小等等)

## 可视化

• 损失函数采用的有问题
• 训练的数据的载入方式可能有问题
• 优化器(optimizer)可能有问题
• 一些其他的超参数设置可能有问题

### 正则化

“Dice” is a a metric for model evaluation equal to intersection(A,B)/(A+B). Similar to IoU (IoU = intersection(A,B)/union(A,B)), it is used to assess a quality of image segmentation models. “Accuracy” is not really good for this task. For example, in this competition, you can quite easily get 99.9% accuracy of predicted pixels, but the performance of the models may be not as great as you think. Meanwhile, such metrics as dice or IoU can describe your models reasonably well, therefore they are most often used to asses publicly available models. The implementation of the metric used for score evaluation in this competition takes some time and requires additional post-processing, such as mask splitting. Therefore, it is not so common for quick model evaluation. Also, sometimes “soft dice” (dice with multiplication instead of intersection) is used as a loss function during training of image segmentation models.

### 标准化和批标准化

batch-normalization的好处：https://www.learnopencv.com/batch-normalization-in-deep-networks/

点赞

欢迎关注Oldpan博客微信公众号，同步更新博客深度学习文章。