如何减少神经网络的过度拟合? [英] how to reduce overfitting in neural networks?
问题描述
我正在做一个声音识别项目.
I'm working on a sound recognition project.
我有1500个带标签的5类声音样本.(每个类别300个声音样本,持续2秒).
I have 1500 labeled sound samples of 5 classes. (300 sound samples of a duration of 2 seconds for each classe).
我正在使用在线工具来计算MFCC系数(Egde脉冲)(因此我无法提供代码),然后我正在训练神经网络.
I'm using an online tool to calculate the MFCC coefficients (Egde impulse) (So I can not provide the code) and then I'm training a neural network.
数据集已拆分:
-
80%->分为80/20的训练集-训练/验证
80% --> a training set which is splitted 80/20 - training/validation
20%->测试集
经过200个训练周期,我的网络的第一个发行版具有以下表现(非常糟糕):
After 200 training cycles, the first release of my network had the (very bad) following performances :
训练准确度= 100%/验证准确度= 30%
training accuracy = 100 % / Validation accuracy = 30 %
通过在网上和这个论坛上搜索,我找到了减少过度拟合的方法:
By searching on the net and on this forum, I found method(s) to reduce overfitting :
我最后一次发布的神经网络的最终性能如下:
The final performance of my last release of neural network is the following :
训练准确度= 80%/验证准确度= 60%(200个训练周期后)
training accuracy = 80 % / Validation accuracy = 60 % (after 200 training cycles)
如您所见,训练准确性和验证准确性之间仍然存在显着差异.
As you can see, there is still a significant difference between training accuracy and validation accuracy..
我的问题是如何继续提高验证准确性?
My question is how to continue to increase my validation accuracy ?
我的神经网络的代码:
import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, InputLayer, Dropout, Conv1D, Flatten, Reshape, MaxPooling1D, BatchNormalization
from tensorflow.keras import regularizers
from tensorflow.keras.optimizers import Adam
# model architecture
model = Sequential()
model.add(InputLayer(input_shape=(X_train.shape[1], ), name='x_input'))
model.add(Reshape((int(X_train.shape[1] / 13), 13), input_shape=(X_train.shape[1], )))
model.add(Conv1D(30, kernel_size=1, activation='relu',kernel_regularizer=regularizers.l2(0.001)))
model.add(Dropout(0.5))
model.add(MaxPooling1D(pool_size=1, padding='same'))
model.add(Conv1D(10, kernel_size=1, activation='relu',kernel_regularizer=regularizers.l2(0.001)))
model.add(Dropout(0.5))
model.add(MaxPooling1D(pool_size=1, padding='same'))
model.add(Flatten())
model.add(Dense(classes, activation='softmax', name='y_pred'))
# this controls the learning rate
opt = Adam(lr=0.005, beta_1=0.9, beta_2=0.999)
#opt = Adadelta(learning_rate=1.0, rho=0.95)
# train the neural network
model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=['accuracy'])
model.fit(X_train, Y_train, batch_size=50, epochs=200, validation_data=(X_test, Y_test), verbose=2)
谢谢
此致
Lionel
推荐答案
通常为减少过度拟合,可以执行以下操作:
In general to reduce overfitting, you can do the following:
- 添加更多正则化功能(例如,辍学率更高的多层辍学)
- 减少功能数量
- 减少网络容量(例如,减少层数或隐藏单元数)
- 减小批次大小
这篇关于如何减少神经网络的过度拟合?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!