超级会员免费看
本文介绍了如何在猫狗图像分类任务中通过数据增强和Dropout策略来减少过拟合。我们首先探讨了数据增强的概念，包括随机旋转、平移等变换。接着将这些变换应用到预处理步骤中，并添加Dropout层到卷积神经网络。经过重新训练，模型的验证准确率显著提高，表明这些策略有效地防止了过拟合。
Cat vs. Dog Image Classification
Exercise 2: Reducing Overfitting
Estimated completion time: 30 minutes
In this notebook we will build on the model we created in Exercise 1 to classify cats vs. dogs, and improve accuracy by employing a couple strategies to reduce overfitting: data augmentation and dropout.
We will follow these steps:
- Explore how data augmentation works by making random transformations to training images.
- Add data augmentation to our data preprocessing.
- Add dropout to the convnet.
- Retrain the model and evaluate loss and accuracy.
Let’s
订阅专栏 解锁全文
扫一扫
481
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
