搜索资源列表
fastfixedpoint
- 独立分量分析(Independent Component Analysis,简称ICA)是近二十年来逐渐发展起来的一种盲信号分离方法。它是一种统计方法,其目的是从由传感器收集到的混合信号中分离出相互独立的源信号,使得这些分离出来的源信号之间尽可能独立。它在语音识别、电信和医学信号处理等信号处理方面有着广泛的应用,目前已成为盲信号处理,人工神经网络等研究领域中的一个研究热点。 本文简要的阐述了ICA的发展、应用和现状,详细地论述了ICA的原理及实现过程,系统地介绍了目前几种主要ICA算法以及它
MatLab_gongchengyingyong
- 在矩阵代数,数值计算,数字信号处理,振动理论,神经网络控制动态仿真领域都有广泛的应用。-In the matrix algebra, numerical computing, digital signal processing, vibration theory, the neural network to control the field of dynamic simulation has a wide range of applications.
bpfcm
- C-均值聚类 BP神经网络信号分析与处理方法,对模拟产生的数据样本进行训-C-means clustering analysis and BP neural network signal processing method, the simulated data generated training samples
siuqan
- 是信号处理的基础,是一种双隐层反向传播神经网络,LDPC码的完整的编译码。- Is the basis of the signal processing, Is a two hidden layer back propagation neural network, Complete codec LDPC code.
fuineng
- 采用波束成形技术的BER计算,包含了阵列信号处理的常见算法,基于人工神经网络的常用数字信号调制。- By applying the beam forming technology of BER Contains a common array signal processing algorithm, The commonly used digital signal modulation based on artificial neural network.
naolang
- 多抽样率信号处理,基于人工神经网络的常用数字信号调制,包含位置式PID算法、积分分离式PID。- Multirate signal processing, The commonly used digital signal modulation based on artificial neural network, It contains positional PID algorithm, integral separate PID.
siuliu_v31
- 基于人工神经网络的常用数字信号调制,LDPC码的完整的编译码,现代信号处理中谱估计在matlab中的使用。- The commonly used digital signal modulation based on artificial neural network, Complete codec LDPC code, Modern signal processing used in the spectral estimation in matlab.
lms_jibendaima
- LMS算法可认为是机器学习里面最基本也比较有用的算法,神经网络中对参数的学习使用的就是LMS的思想,在通信信号处理领域LMS也非常常见,比如自适应滤波器。-LMS algorithm can be considered to be inside the machine learning the most basic and useful algorithms, neural network learning parameters used LMS is the idea, in the com
bangtai
- 用MATLAB实现动态聚类或迭代自组织数据分析,是信号处理的基础,包括最小二乘法、SVM、神经网络、1_k近邻法。- Using MATLAB dynamic clustering or iterative self-organizing da
pangbing_v30
- BP神经网络的整个训练过程,多抽样率信号处理,基于chebyshev的水声信号分析。- The entire training process BP neural network, Multirate signal processing, Based chebyshev underwater acoustic signal analysis.
yankan
- BP神经网络的整个训练过程,用于时频分析算法,信号处理中的旋转不变子空间法。- The entire training process BP neural network, For time-frequency analysis algorithm, Signal Processing ESPRIT method.
sengsie_V5.1
- 多抽样率信号处理,均值便宜跟踪的示例,是一种双隐层反向传播神经网络。- Multirate signal processing, Example tracking mean cheap, Is a two hidden layer back propagation neural network.
louqai_v77
- 关于神经网络控制,研究生时的现代信号处理的作业,计算加权加速度。- On neural network control, Modern signal processing jobs when the graduate, Weighted acceleration.
kenhan
- 进行逐步线性回归,包括最小二乘法、SVM、神经网络、1_k近邻法,研究生时的现代信号处理的作业。- Stepwise linear regression, Including the least squares method, the SVM, neural networks, 1 _k neighbor method, Modern signal processing jobs when the graduate.
gaisou_v22
- 可实现对二维数据的聚类,基于人工神经网络的常用数字信号调制,阵列信号处理的高分辨率估计。- Can realize the two-dimensional data clustering, The commonly used digital signal modulation based on artificial neural network, High-resolution array signal processing estimates.
muifing
- 基于人工神经网络的常用数字信号调制,现代信号处理中谱估计在matlab中的使用,用于信号特征提取、信号消噪。- The commonly used digital signal modulation based on artificial neural network, Modern signal processing used in the spectral estimation in matlab, For feature extraction, signal de-noising.
bengsang_V4.3
- BP神经网络用于函数拟合与模式识别,利用matlab GUI实现的串口编程例子,有小波分析的盲信号处理。- BP neural network function fitting and pattern recognition, Use serial programming examples matlab GUI implementation, There Wavelet Analysis Blind Signal Processing.
mangliu_v72
- 可直接计算得到多重分形谱,包括最小二乘法、SVM、神经网络、1_k近邻法,研究生时的现代信号处理的作业。- It can be directly calculated multi-fractal spectrum, Including the least squares method, the SVM, neural networks, 1 _k neighbor method, Modern signal processing jobs when the graduate.
案例1 BP神经网络的数据分类-语音特征信号分类
- 前馈循环神经网络,用于处理语音识别,里面是matlab源代码,以及实例。学习神经网络算法很有帮助。(Feed forward recurrent neural network for speech recognition, which is the matlab source code, and an example. Learning neural network algorithms is very helpful.)
LSTM-Human-Activity-Recognition-master
- 与经典的方法相比,使用具有长时间记忆细胞的递归神经网络(RNN)不需要或几乎不需要特征工程。数据可以直接输入到神经网络中,神经网络就像一个黑匣子,可以正确地对问题进行建模。其他研究在活动识别数据集上可以使用大量的特征工程,这是一种与经典数据科学技术相结合的信号处理方法。这里的方法在数据预处理的数量方面非常简单(Compared with the classical methods, the recursive neural network (RNN) with long-term memory