【25】ReLU 家族评比 个别使用ReLU LeakyReLU PReLU SELU 来训练

Colab连结

今天要来实验不同的 ReLU 家族的评比，挑战者有

  1. 基本 ReLU
  2. 尾巴翘起来的 Leaky ReLU
  3. 带有参数控制尾巴的 PReLU
  4. 比较新潮的 SELU

说真的，有那麽多 ReLU，但我自己实务上却只使用过第一种基本款，所以我也蛮好奇这次实验会有多大的对比。

而因为这次想跑多次的 epochs 来进行实验，所以资料集的部分改用 beans 来实验。

https://ithelp.ithome.com.tw/upload/images/20211009/20107299ueYmJLWDES.png

实验一：基本 ReLU

def bottleneck(net, filters, out_ch, strides, shortcut=True, zero_pad=False):

    padding = 'valid' if zero_pad else 'same'
    shortcut_net = net

    net = tf.keras.layers.Conv2D(filters * 6, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.ReLU(max_value=6)(net)
    if zero_pad:
        net = tf.keras.layers.ZeroPadding2D(padding=((0, 1), (0, 1)))(net)

    net = tf.keras.layers.DepthwiseConv2D(3, strides=strides, use_bias=False, padding=padding)(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.ReLU(max_value=6)(net)

    net = tf.keras.layers.Conv2D(out_ch, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)

    if shortcut:
        net = tf.keras.layers.Add()([net, shortcut_net])

    return net

def get_mobilenetV2_relu(shape):
    input_node = tf.keras.layers.Input(shape=shape)

    net = tf.keras.layers.Conv2D(32, 3, (2, 2), use_bias=False, padding='same')(input_node)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.ReLU(max_value=6)(net)

    net = tf.keras.layers.DepthwiseConv2D(3, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.ReLU(max_value=6)(net)
    net = tf.keras.layers.Conv2D(16, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)

    net = bottleneck(net, 16, 24, (2, 2), shortcut=False, zero_pad=True)  # block_1
    net = bottleneck(net, 24, 24, (1, 1), shortcut=True)  # block_2

    net = bottleneck(net, 24, 32, (2, 2), shortcut=False, zero_pad=True)  # block_3
    net = bottleneck(net, 32, 32, (1, 1), shortcut=True)  # block_4
    net = bottleneck(net, 32, 32, (1, 1), shortcut=True)  # block_5

    net = bottleneck(net, 32, 64, (2, 2), shortcut=False, zero_pad=True)  # block_6
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_7
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_8
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_9

    net = bottleneck(net, 64, 96, (1, 1), shortcut=False)  # block_10
    net = bottleneck(net, 96, 96, (1, 1), shortcut=True)  # block_11
    net = bottleneck(net, 96, 96, (1, 1), shortcut=True)  # block_12

    net = bottleneck(net, 96, 160, (2, 2), shortcut=False, zero_pad=True)  # block_13
    net = bottleneck(net, 160, 160, (1, 1), shortcut=True)  # block_14
    net = bottleneck(net, 160, 160, (1, 1), shortcut=True)  # block_15

    net = bottleneck(net, 160, 320, (1, 1), shortcut=False)  # block_16

    net = tf.keras.layers.Conv2D(1280, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.ReLU(max_value=6)(net)

    return input_node, net

input_node, net = get_mobilenetV2_relu((224,224,3))
net = tf.keras.layers.GlobalAveragePooling2D()(net)
net = tf.keras.layers.Dense(NUM_OF_CLASS)(net)

model = tf.keras.Model(inputs=[input_node], outputs=[net])

model.compile(
    optimizer=tf.keras.optimizers.SGD(LR),
    loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
    metrics=[tf.keras.metrics.SparseCategoricalAccuracy()],
)

history = model.fit(
    ds_train,
    epochs=EPOCHS,
    validation_data=ds_test,
    verbose=True)

产出：

Epoch 96/100
loss: 0.0074 - sparse_categorical_accuracy: 0.9961 - val_loss: 0.9123 - val_sparse_categorical_accuracy: 0.8516

https://ithelp.ithome.com.tw/upload/images/20211009/20107299JnJY2sBwi8.png

实验二：Leaky ReLU

一般的 ReLU 遇到输入为负值时，输出会变成0，很容易让大半的节点死亡，因此我们在输入负值时，仍然给它一个 α 值，让他微分时仍然可以更新权重，在这边我们用预设的0.3。

def bottleneck(net, filters, out_ch, strides, shortcut=True, zero_pad=False):

    padding = 'valid' if zero_pad else 'same'
    shortcut_net = net

    net = tf.keras.layers.Conv2D(filters * 6, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.LeakyReLU()(net)
    if zero_pad:
        net = tf.keras.layers.ZeroPadding2D(padding=((0, 1), (0, 1)))(net)

    net = tf.keras.layers.DepthwiseConv2D(3, strides=strides, use_bias=False, padding=padding)(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.LeakyReLU()(net)

    net = tf.keras.layers.Conv2D(out_ch, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)

    if shortcut:
        net = tf.keras.layers.Add()([net, shortcut_net])

    return net

def get_mobilenetV2_leakyrelu(shape):
    input_node = tf.keras.layers.Input(shape=shape)

    net = tf.keras.layers.Conv2D(32, 3, (2, 2), use_bias=False, padding='same')(input_node)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.LeakyReLU()(net)

    net = tf.keras.layers.DepthwiseConv2D(3, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.LeakyReLU()(net)
    net = tf.keras.layers.Conv2D(16, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)

    net = bottleneck(net, 16, 24, (2, 2), shortcut=False, zero_pad=True)  # block_1
    net = bottleneck(net, 24, 24, (1, 1), shortcut=True)  # block_2

    net = bottleneck(net, 24, 32, (2, 2), shortcut=False, zero_pad=True)  # block_3
    net = bottleneck(net, 32, 32, (1, 1), shortcut=True)  # block_4
    net = bottleneck(net, 32, 32, (1, 1), shortcut=True)  # block_5

    net = bottleneck(net, 32, 64, (2, 2), shortcut=False, zero_pad=True)  # block_6
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_7
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_8
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_9

    net = bottleneck(net, 64, 96, (1, 1), shortcut=False)  # block_10
    net = bottleneck(net, 96, 96, (1, 1), shortcut=True)  # block_11
    net = bottleneck(net, 96, 96, (1, 1), shortcut=True)  # block_12

    net = bottleneck(net, 96, 160, (2, 2), shortcut=False, zero_pad=True)  # block_13
    net = bottleneck(net, 160, 160, (1, 1), shortcut=True)  # block_14
    net = bottleneck(net, 160, 160, (1, 1), shortcut=True)  # block_15

    net = bottleneck(net, 160, 320, (1, 1), shortcut=False)  # block_16

    net = tf.keras.layers.Conv2D(1280, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.LeakyReLU()(net)

    return input_node, net

input_node, net = get_mobilenetV2_leakyrelu((224,224,3))
net = tf.keras.layers.GlobalAveragePooling2D()(net)
net = tf.keras.layers.Dense(NUM_OF_CLASS)(net)

model = tf.keras.Model(inputs=[input_node], outputs=[net])

model.compile(
    optimizer=tf.keras.optimizers.SGD(LR),
    loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
    metrics=[tf.keras.metrics.SparseCategoricalAccuracy()],
)

history = model.fit(
    ds_train,
    epochs=EPOCHS,
    validation_data=ds_test,
    verbose=True)

产出:

Epoch 96/100
loss: 0.0074 - sparse_categorical_accuracy: 0.9981 - val_loss: 1.2042 - val_sparse_categorical_accuracy: 0.7969

https://ithelp.ithome.com.tw/upload/images/20211009/20107299jFomliOjmU.png

实验三：PReLU

有了 LeakyReLU 後，就会有人觉得给予固定的 α 值似乎让模型缺乏了弹性，因此不如让这个值也变成模型学习的参数，所以就有了 PReLU，在这边我们让他的初始值为0.25。

def bottleneck(net, filters, out_ch, strides, shortcut=True, zero_pad=False):

    padding = 'valid' if zero_pad else 'same'
    shortcut_net = net

    net = tf.keras.layers.Conv2D(filters * 6, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.PReLU(alpha_initializer=tf.initializers.constant(0.25))(net)
    if zero_pad:
        net = tf.keras.layers.ZeroPadding2D(padding=((0, 1), (0, 1)))(net)

    net = tf.keras.layers.DepthwiseConv2D(3, strides=strides, use_bias=False, padding=padding)(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.PReLU(alpha_initializer=tf.initializers.constant(0.25))(net)

    net = tf.keras.layers.Conv2D(out_ch, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)

    if shortcut:
        net = tf.keras.layers.Add()([net, shortcut_net])

    return net

def get_mobilenetV2_prelu(shape):
    input_node = tf.keras.layers.Input(shape=shape)

    net = tf.keras.layers.Conv2D(32, 3, (2, 2), use_bias=False, padding='same')(input_node)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.PReLU(alpha_initializer=tf.initializers.constant(0.25))(net)

    net = tf.keras.layers.DepthwiseConv2D(3, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.PReLU(alpha_initializer=tf.initializers.constant(0.25))(net)
    net = tf.keras.layers.Conv2D(16, 1, use_bias=False, padding='same')(net)
    net = tf.keras.layers.BatchNormalization()(net)

    net = bottleneck(net, 16, 24, (2, 2), shortcut=False, zero_pad=True)  # block_1
    net = bottleneck(net, 24, 24, (1, 1), shortcut=True)  # block_2

    net = bottleneck(net, 24, 32, (2, 2), shortcut=False, zero_pad=True)  # block_3
    net = bottleneck(net, 32, 32, (1, 1), shortcut=True)  # block_4
    net = bottleneck(net, 32, 32, (1, 1), shortcut=True)  # block_5

    net = bottleneck(net, 32, 64, (2, 2), shortcut=False, zero_pad=True)  # block_6
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_7
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_8
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_9

    net = bottleneck(net, 64, 96, (1, 1), shortcut=False)  # block_10
    net = bottleneck(net, 96, 96, (1, 1), shortcut=True)  # block_11
    net = bottleneck(net, 96, 96, (1, 1), shortcut=True)  # block_12

    net = bottleneck(net, 96, 160, (2, 2), shortcut=False, zero_pad=True)  # block_13
    net = bottleneck(net, 160, 160, (1, 1), shortcut=True)  # block_14
    net = bottleneck(net, 160, 160, (1, 1), shortcut=True)  # block_15

    net = bottleneck(net, 160, 320, (1, 1), shortcut=False)  # block_16

    net = tf.keras.layers.Conv2D(1280, 1, use_bias=False, padding='same',)(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.layers.PReLU(alpha_initializer=tf.initializers.constant(0.25))(net)

    return input_node, net

input_node, net = get_mobilenetV2_prelu((224,224,3))
net = tf.keras.layers.GlobalAveragePooling2D()(net)
net = tf.keras.layers.Dense(NUM_OF_CLASS)(net)

model = tf.keras.Model(inputs=[input_node], outputs=[net])

model.compile(
    optimizer=tf.keras.optimizers.SGD(LR),
    loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
    metrics=[tf.keras.metrics.SparseCategoricalAccuracy()],
)

history = model.fit(
    ds_train,
    epochs=EPOCHS,
    validation_data=ds_test,
    verbose=True)

产出:

Epoch 99/100
loss: 0.0119 - sparse_categorical_accuracy: 0.9961 - val_loss: 1.3487 - val_sparse_categorical_accuracy: 0.8125

https://ithelp.ithome.com.tw/upload/images/20211009/20107299wmeO9gzCR6.png

实验四：SELU

介绍 SELU 之前，要先知道他是 ELU 的一种，ELU在输入小於0时，会是一个 exponential 形状，所以在计算量上都会比较大，ELU 的公式如下：

https://ithelp.ithome.com.tw/upload/images/20211009/201072996o0SS6YHQP.png

https://ithelp.ithome.com.tw/upload/images/20211009/20107299SrIOOoeSFV.png
来源

而 SELU 公式和 ELU 差不多，只是原作者找到了一个确切的 α 数值和一个固定的 λ 值来做缩放，而且在使用这个 act func 时，前面的权重初始化要使用 lecun_normal，如果有用过 dropout 则是要替换成另一种 AlphaDropout。

def bottleneck(net, filters, out_ch, strides, shortcut=True, zero_pad=False):

    padding = 'valid' if zero_pad else 'same'
    shortcut_net = net

    net = tf.keras.layers.Conv2D(filters * 6, 1, use_bias=False, padding='same', kernel_initializer='lecun_normal')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.activations.selu(net)
    if zero_pad:
        net = tf.keras.layers.ZeroPadding2D(padding=((0, 1), (0, 1)))(net)

    net = tf.keras.layers.DepthwiseConv2D(3, strides=strides, use_bias=False, padding=padding, kernel_initializer='lecun_normal')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.activations.selu(net)

    net = tf.keras.layers.Conv2D(out_ch, 1, use_bias=False, padding='same', kernel_initializer='lecun_normal')(net)
    net = tf.keras.layers.BatchNormalization()(net)

    if shortcut:
        net = tf.keras.layers.Add()([net, shortcut_net])

    return net

def get_mobilenetV2_selu(shape):
    input_node = tf.keras.layers.Input(shape=shape)

    net = tf.keras.layers.Conv2D(32, 3, (2, 2), use_bias=False, padding='same', kernel_initializer='lecun_normal')(input_node)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.activations.selu(net)

    net = tf.keras.layers.DepthwiseConv2D(3, use_bias=False, padding='same', kernel_initializer='lecun_normal')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.activations.selu(net)
    net = tf.keras.layers.Conv2D(16, 1, use_bias=False, padding='same', kernel_initializer='lecun_normal')(net)
    net = tf.keras.layers.BatchNormalization()(net)

    net = bottleneck(net, 16, 24, (2, 2), shortcut=False, zero_pad=True)  # block_1
    net = bottleneck(net, 24, 24, (1, 1), shortcut=True)  # block_2

    net = bottleneck(net, 24, 32, (2, 2), shortcut=False, zero_pad=True)  # block_3
    net = bottleneck(net, 32, 32, (1, 1), shortcut=True)  # block_4
    net = bottleneck(net, 32, 32, (1, 1), shortcut=True)  # block_5

    net = bottleneck(net, 32, 64, (2, 2), shortcut=False, zero_pad=True)  # block_6
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_7
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_8
    net = bottleneck(net, 64, 64, (1, 1), shortcut=True)  # block_9

    net = bottleneck(net, 64, 96, (1, 1), shortcut=False)  # block_10
    net = bottleneck(net, 96, 96, (1, 1), shortcut=True)  # block_11
    net = bottleneck(net, 96, 96, (1, 1), shortcut=True)  # block_12

    net = bottleneck(net, 96, 160, (2, 2), shortcut=False, zero_pad=True)  # block_13
    net = bottleneck(net, 160, 160, (1, 1), shortcut=True)  # block_14
    net = bottleneck(net, 160, 160, (1, 1), shortcut=True)  # block_15

    net = bottleneck(net, 160, 320, (1, 1), shortcut=False)  # block_16

    net = tf.keras.layers.Conv2D(1280, 1, use_bias=False, padding='same', kernel_initializer='lecun_normal')(net)
    net = tf.keras.layers.BatchNormalization()(net)
    net = tf.keras.activations.selu(net)

    return input_node, net

input_node, net = get_mobilenetV2_selu((224,224,3))
net = tf.keras.layers.GlobalAveragePooling2D()(net)
net = tf.keras.layers.Dense(NUM_OF_CLASS)(net)

model = tf.keras.Model(inputs=[input_node], outputs=[net])

model.compile(
    optimizer=tf.keras.optimizers.SGD(LR),
    loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
    metrics=[tf.keras.metrics.SparseCategoricalAccuracy()],
)

history = model.fit(
    ds_train,
    epochs=EPOCHS,
    validation_data=ds_test,
    verbose=True)

产出：

Epoch 85/100
loss: 0.0018 - sparse_categorical_accuracy: 1.0000 - val_loss: 0.6651 - val_sparse_categorical_accuracy: 0.8672

https://ithelp.ithome.com.tw/upload/images/20211009/20107299cQzVQSjem3.png

以上就是今天针对这四个不同的 ReLU 所做的实验，大致上都有得到不错的效果，今天如果你要处理的任务比较急或者是第一次训练的话，或许还是基本款的 ReLU 就好，如果是相比 Kaggle 这种必须计较分数的比赛，那就很推荐尝试一些新的手法来训练。


<<:  Day 25 - 将 Yacht Manager 後台储存资料提取後，送至前台渲染 Yachts 版面重覆区块 - 前台 Master Page 制作 - ASP.NET Web Forms C#

>>:  如何在 Angular 获取 URL 资讯

Day.24 「你点了按钮～同时也点了网页本身！」 —— JavaScript 事件冒泡（Event bubbling）

我们上一个篇章认识了绑定事件，了解到不管是什麽节点，都可以绑定事件 那为什麽 JavaScript...

D16/ 所以到底为什麽 remember 是 composable function? - @Composable 是什麽 part 2

今天大概会聊到的范围 compose runtime compose compiler 今天会更深...

微聊 术业有专攻 Blame要找对人

倒数第二天铁人赛了，微人今日不聊程序码了，想聊一下今天在预演的小插曲 先说明一下 我的队友们真的都很...

[重构倒数第11天] - 如何在 Vue 中写出高效能的网页渲染方式 ?

前言 该系列是为了让看过Vue官方文件或学过Vue但是却不知道怎麽下手去重构现在有的网站而去规画的系...

【把玩Azure DevOps】Day12 Artifacts应用：上传第一个nuget package

前一篇文章简单介绍了Azure DevOps Artifacts，知道了它就是用来存放私有套件的套件...