搜索资源列表
Narama-L2
- 介绍基于神经网络的反馈线性化控制过程。反馈线性就是利用反馈的控制手段来消除系统中的非线性,以使的其闭环系统的动力学方程是线性的。-Introduced based on neural network feedback linearization control process. Feedback linearization is to use feedback control to eliminate the non-linear system, so that its closed-loop
libsvm-2.89
- 是一種線性方成的分類器。SVM透過統計的方式將雜亂的資料以NN的方式分成兩類,以便處理。LIBLINEAR is a linear classifier for data with millions of instances and features. It supports L2-regularized logistic regression (LR), L2-loss linear SVM, and L1-loss linear SVM. -Main features of LIBLINEA
Sparse-Reconstruction_2.0
- 基于分散近似的稀疏重建算法,用于l1-l2规划,图像去噪,图像压缩等。-sparse reconstruction by separable approximation
Neural-network-quxianbijin
- 设计并训练三种神经网络使之分别逼近下列函数,精度Sm偏差小于20,Ry偏差小于1.5。 (各变量取值范围: =20~90, =35~55,a1=3~13, a2=0.3~3,Sd=0.05~0.45, =0.05~0.04,L2=0.015~0.06,T=7~110)-Design and training of three neural networks respectively approximation of the following function, precision Sm dev
GPS-cycle-slip-processing
- GPS L1,L2双拼周跳探测与修复,包含几种常用的实时探测算法,附有GUI界面-A MATLAB software package for GPS cycle-slip processing is presented in this paper. It realizes cycle-slip detection and repair in the measurement domain for GPS L1 and L2 signals. The software implement
bpback
- 神经网络比较基础的算法,实现梯度下降和反向传播,以及L2规范化、交叉熵代价函数的引入,卷积神经网络 该算法用于mnist数据测试,有详细中文注释-Neural network based on the comparison algorithm, gradient descent and back-propagation, and L2 standardization introduced cross entropy cost function, convolution neural netw
bsvm-2.08
- BSVM解决了支持向量机(SVM),用于解决大型分类和回归问题。 它包括以下方法 一个对一个使用约束约束公式的多类分类 通过解决单一优化问题(再次,有界公式)进行多类分类。 参见我们比较文件的第3节。 使用Crammer和Singer的配方进行多级分类。 参见我们的比较文章第4节。 使用约束约束公式的回归-BSVM solves support vector machines (SVM) for the solution of large classification and r
正则化
- 深度网络中防至过拟合的情况,可以使用正则化,这里将实现L2正则化(In case of overfitting in deep network, regularization can be used, and L2 regularization is implemented here.)