简体繁体 English

Keras 迁移学习

[英]Keras Transfer Learning

原文 2020-01-28 10:41:27 8 1 python/ machine-learning/ keras/ conv-neural-network/ transfer-learning

Let's say that I wanted to train a ConvNet to tell me if in a photo it is raining or not, how will the procedures be?假设我想训练一个 ConvNet 来告诉我照片中是否下雨，程序将如何？

Given that I have two train variables, trainX and trainY , the trainX will be the photo and trainY will be the labels (eg rain or no-rain ).鉴于我有两个火车变量， trainX和trainY ， trainX将是照片， trainY将是标签（例如rain或no-rain ）。

The goal of the network is to output the "right" answer.网络的目标是输出“正确”的答案。 The question is: do I just need to run the model.predict() function and expect valid results?问题是：我是否只需要运行model.predict()函数并期望得到有效结果？

Thank you for any help in advance.提前感谢您的任何帮助。

1 个解决方案

1) Build your CNN Model: Layers, Activation-Functions... 1) 构建您的 CNN 模型：层、激活函数...

2) Train it with your existing trainX and trainY -Dataset. 2) 使用您现有的trainX和trainY -Dataset trainY训练。 (use Augmentation to get better results in the end) （最终使用Augmentation获得更好的结果）

3) Validate with another Dataset, lets say they are called: testX and testY 3) 使用另一个数据集进行验证，假设它们被称为： testX和testY

4) Modify the settings of your Model until your accuracy and loss are high enough for what you need them... 4）修改模型的设置，直到您的准确度和损失足够高以满足您的需要...

5) enjoy your CNN 5）享受你的CNN

This could help you on your way: Building a CNN with Keras这可以为您提供帮助：使用 Keras 构建 CNN