小编Ren*_*enz的帖子

我使用卷积神经网络(通过Keras)作为我的面部表情识别模型(55名受试者).我的数据集非常难,大约450k,有7个班级.我根据每个科目和每个班级标签平衡了我的训练集.

我实现了一个非常简单的CNN架构(具有实时数据增强功能):

model = Sequential()
model.add(Convolution2D(32, 3, 3, border_mode=borderMode, init=initialization,  input_shape=(48, 48, 3)))
model.add(BatchNormalization())
model.add(PReLU())
model.add(MaxPooling2D(pool_size=(2, 2)))

model.add(Flatten())
model.add(Dense(256))
model.add(BatchNormalization())
model.add(PReLU())
model.add(Dropout(0.5))

model.add(Dense(nb_output))
model.add(Activation('softmax'))

在第一个时期之后,我的训练损失不断减少,而验证损失增加.很快会过度拟合吗？或者我的数据是否存在混淆问题？我还应该平衡我的测试集吗？

我按照如何安装Laravel,但毕竟当我进入"vagrant up"时我得到这些错误:

C:/Users/Renz/code/Homestead/scripts/homestead.rb:63:在configure': undefined m ethod每个'for nil:NilClass(NoMethodError)来自C:/ Users/Renz/code/Homestead/Vagrantfile:9:in call'from block in <top (requi red)>' from C:/HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/lib/vagrant/c onfig/v2/loader.rb:37:inC :load' from C:/HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/lib/vagrant/c onfig/loader.rb:103:in/ HashiCorp/Vagrant/embedded/gems/ gems/vagrant-1.6.5/lib/vagrant/c onfig/v2/loader.rb:37:in block(2 level)in load'from C:/ HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/lib/vagrant/c onfig/loader.rb:97:in in each' from C:/HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/lib/vagrant/c onfig/loader.rb:97:inblock in load'from C:/HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/ lib/vagrant/c onfig/loader.rb:94:in each' from C:/HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/lib/vagrant/c onfig/loader.rb:94:inload'from C:/HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/lib/vagrant/v agrantfile.rb:28:in initialize' from C:/HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/lib/vagrant/e nvironment.rb:614:innew'来自C:/HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/lib/vagrant/e nvironment.rb:614:在vagrantfile' from C:/HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/lib/vagrant/e nvironment.rb:399:in主机'来自C:/ HashiCorp/Vagrant/embedded/gems/gems/vagrant- 1.6.5/lib/vagrant/e nvironment.rb:198:block in action_runner' from C:/HashiCorp/Vagrant/embedded/gems/gems/vagrant-1.6.5/lib/vagrant/a ction/runner.rb:33:in来自C的电话:/ HashiCorp/Vagrant/embedded/gems/gems /vagrant-1.6.5/lib/vagrant/ a ction/runner.rb:33 :run' …