Extreme learning machine in study on application o
2016-08-23
3 0 0
4.0
Other
Earn points
Single hidden-layer feedforward neural networks (Single-hidden Layer Feedforward Neural Network,SLFN) for its excellent ability to learn has been widely applied in many fields. However, the conventional learning algorithms (such as BP algorithm and so on) some inherent disadvantages become the bottleneck of the main constraints to their development. Most of feedforward neural networks using gradient descent method in this method there are several shortcomings and deficiencies:
(1) training speed. Due to the gradient descent method requires multiple iterations, so as to achieve the purpose of amending weights and thresholds, so the training process takes a long time;
(2) easily into local minima, unable to reach the global minimum;
(3) the learning rate-sensitive. Great influence on the performance of neural network learning rate, you must choose the right to get an ideal network. If it is too small, then the convergence is slow, training process takes too long; Conversel
(1) training speed. Due to the gradient descent method requires multiple iterations, so as to achieve the purpose of amending weights and thresholds, so the training process takes a long time;
(2) easily into local minima, unable to reach the global minimum;
(3) the learning rate-sensitive. Great influence on the performance of neural network learning rate, you must choose the right to get an ideal network. If it is too small, then the convergence is slow, training process takes too long; Conversel
matlab
分类
回归
应用
研究
拟合
实验
问题
对比
极限
学习机
Related Source Codes
GMSK Linear Receiver
0
0
no vote
NSGA-II algorithm
0
0
no vote
NSGA-III multi-objective optimization algorithm
0
0
no vote
Compressed sensing example
0
0
no vote
CFAR detector example
0
0
no vote
No comment