求助：一审结果，评审人的问题不知怎么回答!!!
我做的深度学习DBN与模糊聚类结合的神经网络算法。
算法用UCI的数据与传统的bp神经网络和DBN进行了对比。验证了算法的性能。
现在审稿人问我：
1. The algorithm is only given by description instead of the theoretical derivation. Then, how is the accuracy guaranteed? What is the rate of success?
2. what is the time complexity of this algorithm ?
我想问一下，神经网络有成功率这个概念么？精度保证？这不是和对象有关的么？
算法时间复杂度这个该怎么回答？是要我数学写出算法时间复杂度么（这个有点难吧）？
返回小木虫查看更多
今日热帖
京公网安备 11010802022153号
https://www.ncku.edu.tw/source/h ... ural%20Networks.pdf
After the training work of the BPNN model is completed, the
performance validation and the model verification are conducted
by an inside test and an outside test. The inside test is used to
check the performance of network learning by using all of the
training data set (Haykin, 2007).
4.1. Inside test
4.2. Outside test，
非常感谢~~
DBN算法复杂度o(n^2)
楼主大神啊，感觉DBN很复杂的，我现在打算将其应用于医学图像方面，楼主用了改进的DBN吗
楼主你的论文已发表了吗