R中的python keras和keras之间的准确性不同 [英] Different accuracy between python keras and keras in R

问题描述

我在keras的R中为R建立了图像分类模型.

I build a image classification model in R by keras for R.

大约有98％的准确性，而python的准确性却很差.

Got about 98% accuracy, while got terrible accuracy in python.

R的Keras版本是2.1.3，而python中是2.1.5

Keras version for R is 2.1.3, and 2.1.5 in python

以下是R模型代码:

model=keras_model_sequential()
model=model %>% 
  layer_conv_2d(filters = 32,kernel_size = c(3,3),padding = 'same',input_shape = c(187,256,3),activation = 'elu')%>%
  layer_max_pooling_2d(pool_size = c(2,2)) %>%
  layer_dropout(.25) %>% layer_batch_normalization() %>%
  layer_conv_2d(filters = 64,kernel_size = c(3,3),padding = 'same',activation = 'relu') %>%
  layer_max_pooling_2d(pool_size = c(2,2)) %>%
  layer_dropout(.25) %>% layer_batch_normalization() %>% layer_flatten() %>%
  layer_dense(128,activation = 'relu') %>%
  layer_dropout(.25)%>%
  layer_batch_normalization() %>%
  layer_dense(6,activation = 'softmax')


model %>%compile(
  loss='categorical_crossentropy',
  optimizer='adam',
  metrics='accuracy'
)

我尝试使用相同的输入数据在python中重建相同的模型.

I try to rebuild a same model in python, with same input data.

而性能却完全不同.准确度甚至不到30％

While, got totally different performance. The accuracy even less than 30%

因为R keras调用python来运行keras.在相同的模型架构下，它们应该获得相似的性能.

Because R keras is calling python for run keras. With same model architecture, they should get similar performance.

我想知道这个问题是否是由预处理引起的，但仍然显示我的python代码:

I wonder if this issue caused by preprocess, but still show my python code:

model=Sequential()
model.add(Conv2D(32,kernel_size=(3,3),activation='relu',input_shape=(187,256,3),padding='same'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(BatchNormalization())
model.add(Conv2D(64, (3, 3), activation='relu',padding='same'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(BatchNormalization())
model.add(Flatten())
model.add(Dense(128, activation='relu'))
model.add(Dropout(0.25))
model.add(BatchNormalization())
model.add(Dense(len(label[1]), activation='softmax'))

model.compile(loss='categorical_crossentropy',
              optimizer='adam',
              metrics=['accuracy'])

这是一个简单的分类.我和大多数指令一样.

This is a simple classification. I just do as same as most instruction.

找不到其他面临相同问题的人.因此想问一下它是如何发生的以及如何解决.谢谢

Can not find others faced same problem. So want to ask how it happen and how to solve. Thx

推荐答案

这是一个巨大的差异，因此也许代码中有错误或数据中有意外内容，但从Python中的R复制了Keras结果由于将种子设置在R一侧是不够的，所以它比看起来要困难的多.代替set.seed，您应该使用use_session_with_seed，它随tensorflow和keras的R库一起提供.请注意，要完全再现，您需要use_session_with_seed(..., disable_gpu=TRUE, disable_parallel_cpu=TRUE).另请参见 stack 和 tf 文档.另外，这是一个示例，它使用kerasformula和一个公共数据集.另外，请注意像layer_dropout这样的接受seed作为参数的函数.

That is a dramatic difference so perhaps there's a bug in the code or something unexpected in the data but reproducing Keras results from R in Python is more difficult than it may seem since setting the seed on the R side is insufficient. Instead of set.seed you should use use_session_with_seed, which comes with the R libraries for tensorflow and keras. Note that for full reproducibility you need to use_session_with_seed(..., disable_gpu=TRUE, disable_parallel_cpu=TRUE). See also stack and tf docs. Also, here is an example using the github version of kerasformula and a public dataset. Also, watch out for functions like layer_dropout that accept seed as a parameter.

这篇关于R中的python keras和keras之间的准确性不同的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！