Description: 本程序实做MLP(Multi-layer perceptron)算法，使用者可以自行设定训练数据集与测试数据集，将训练数据集加载，在2、3维下可以显示其分布状态，并分别设定键节值、学习率、迭代次数来训练其类神经网络，最后可观看辨识率与RMSE（Root Mean squared error）来判别训练是否可以停止。-This procedure is to do MLP (Multi-layer perceptron) algorithm, the user can set their own training data set and test data sets, the training data set is loaded, in the 2,3-dimensional display of their distribution, and were set key section of the value of learning rate, number of iterations to train the neural network can watch the final recognition rate and the RMSE (Root Mean squared error) to determine whether the training can stop. Platform: |
Size: 1213440 |
Author:楊易 |
Hits:
Description: This demo nstrates how to use the sequential Monte Carlo algorithm with reversible jump MCMC steps to perform model selection in neural networks. We treat both the model dimension (number of neurons) and model parameters as unknowns. The derivation and details are presented in: Christophe Andrieu, Nando de Freitas and Arnaud Doucet. Sequential Bayesian Estimation and Model Selection Applied to Neural Networks . Technical report CUED/F-INFENG/TR 341, Cambridge University Department of Engineering, June 1999. After downloading the file, type "tar -xf version2.tar" to uncompress it. This creates the directory version2 containing the required m files. Go to this directory, load matlab5 and type "smcdemo1". In the header of the demo file, one can select to monitor the simulation progress (with par.doPlot=1) and modify the simulation parameters.-This demo nstrates how to use the sequential Monte Carlo algorithm with reversible jump MCMC steps to perform model selection in neural networks. We treat both the model dimension (number of neurons) and model parameters as unknowns. The derivation and details are presented in: Christophe Andrieu, Nando de Freitas and Arnaud Doucet. Sequential Bayesian Estimation and Model Selection Applied to Neural Networks . Technical report CUED/F-INFENG/TR 341, Cambridge University Department of Engineering, June 1999. After downloading the file, type "tar-xf version2.tar" to uncompress it. This creates the directory version2 containing the required m files. Go to this directory, load matlab5 and type "smcdemo1". In the header of the demo file, one can select to monitor the simulation progress (with par.doPlot=1) and modify the simulation parameters. Platform: |
Size: 220160 |
Author:晨间 |
Hits:
Description: 提出了一种基于函数联接的感知器神经网络的纹理分类方法.它采用高斯2马尔柯夫随机场模型(GM RF)对纹理进行描述,模型参数即为纹理特征,参数估计采用最小平方误差方法获得.将估计参数作为表达纹理的特征向量,用感知器网络对特征进行分类,并且采用函数联接的方式解决线性不可分问题.对纹理图象进行的实验表明,采用这种方法能够提高学习速度,简化计算过程,并取得较好的纹理分类效果.
-Based on the function connected perceptron neural network texture classification method. It uses 2 Gaussian Markov Random Field Model (GM RF) to describe the texture, the model parameters is the texture feature, parameter estimation using least squares error obtained. the estimated parameters as the expression of texture feature vector, using the characteristics of sensor networks for classification, and the use of function to resolve connection problems can not be separated from linear. of texture images of the experiments show that this approach can enhance the learning speed, to simplify the calculation process and obtain a better effect of texture classification. Platform: |
Size: 285696 |
Author:singro jiang |
Hits:
Description: this MLP project in Neural Network that have userinterface. run GUI.m to execute project -this is MLP project in Neural Network that have userinterface. run GUI.m to execute project Platform: |
Size: 24576 |
Author:autstudent |
Hits:
Description: 很全的数字信号C语言资料，包括：数字信号的产生，数字信号处理（FFT计算等），快速离散正交变换，快速卷积与相关，数字滤波器的时域和频域相应，IIR数字滤波器的设计，FIR数字滤波器的设计，以及谱估计，数字图像处理，人工神经网络方面的内容-The digital signal is full C language, including: the emergence of digital signal, digital signal processing (FFT calculation, etc.), fast discrete orthogonal transform, fast convolution and correlation, digital filter in time domain and corresponding frequency domain, IIR digital filter design, FIR digital filter design, as well as the spectral estimation, digital image processing, artificial neural network aspects Platform: |
Size: 6790144 |
Author:李元柳 |
Hits:
Description: 在机动目标跟踪中，机动目标模型是机动目标跟踪的基本要素之一，一般希望机动目标模型能准确表征目标机动时的各种运动状态。比较常用的模型有匀速运动(CV)模型、匀加速运动(CA) 模型、时间相关模型（Singer）和机动目标“当前”统计模型。上述模型均采用机动频率表征目标的机动情况。在应用当中，通常采用固定的机动频率，这就表示机动目标的机动时间是一定的，而实际上机动目标的机动时间是不断变化的，也就是说机动频率是不断变化的，采用固定机动频率必然会带来误差。采样周期在0.5—2S时，机动频率越小跟踪精度越高[1]，但机动频率仍然是固定值。本文提出的基于神经网络的机动频率自适应调整方法可以使机动频率随机动而变化，从而提高状态估计的准确性，提高跟踪精度。本文将小波神经网络用于机动目标跟踪中机动频率的自适应调整，该算法对机动目标“当前”统计模型中的机动频率进行实时修改, 从而自适应的改变机动频率,使跟踪算法与目标的真实状态更接近。该算法采用小波神经网络的离线训练，实时性好。-The maneuver of the maneuvering target is uncertain. The maneuvering frequency is constantly changeable, but traditionally it is beforehand determined as a constant based on the target state estimation in the state model of the maneuvering target. The maneuver of the maneuvering target makes the kinematics equation of the target model mismatch with the practical motion model and the tracking error will be increased. Based on the advantages of the self-learning, the rapid convergence rate and the nonlinear approximation ability of the wavelet neural network, it was put forward to be used in the field of target tracking in the paper. The new residual is used as the input of the wavelet neural network, the output of the network is used to adjust adaptively the maneuvering frequency of the CS model. The algorithm is more close to the real state of the target. The simulation results showed that tracking error can be reduced and the tracking accuracy can be improved. Platform: |
Size: 4096 |
Author:李隆基 |
Hits:
Description: Developing Models from Experimental Data using System Identification Toolbox-1. webinar_walk_through.m: contains all the linear and nonlinear estimation examples presented during the webinar.
2. Data files and Simulink models: process_data.mat, ExampleModel.mdl, Friction_Model.mdl. Any other data files used in the presentation already ship with the toolbox (ver 7.0).
Products used:
- You basically need only System Identification Toolbox (SITB) to try out most examples.
- To use Simulink blocks, you would, of course, need Simulink.
- Control System Toolbox is used at one place to show how estimated models can be converted into LTI objects (SS, TF etc)
- Optimization Toolbox will be used if available for grey box estimation. If not, SITB s built-in optimizers will be used automatically.
- Other products mentioned: Neural Network Toolbox, Model Predictive Control Toolbox and Robust Control Toolbox.
Platform: |
Size: 34816 |
Author:陈翼男 |
Hits:
Description: 神经网络bp算法VC++实现网络的相关运算有：1、网络的输入输出接口，即训练数据的输入，各层权值和节点阈值的输出；2、网络的学习，包括前向传播运算和反向传播运算，误差估计，权值阈值修改；3、网络预测的实现等等。其中网络的学习算法采用变步长和加动量项的优化学习算法，经过我的实验对网络的学习效率有很大提高-Neural network bp algorithm VC++ to achieve the network-related operations: 1, the network input and output interfaces, that is, training data input, each layer weights and node thresholds output 2, the network' s learning, including prior to the spread of computing and the anti- to the spread of computing, error estimation, the right to modify the threshold value 3, the realization of the network prediction and so on. Which the network learning algorithm using variable step size and processing optimization algorithm with momentum term, after my experiment, network efficiency has greatly improved Platform: |
Size: 254976 |
Author:dcw |
Hits:
Description: This paper describes a Model Reference Adaptive
System (MRAS) based scheme using a multilayer
Recurrent Neural Network (RNN) for online speed
estimation of sensorless vector controlled inductmon
motor drive.
Platform: |
Size: 11264 |
Author:chinni |
Hits:
Description: 著名的卡尔曼滤波器，植根于状态空间线性动力学系统，对线性递归最优滤波问题提出了一个叠代的解决方法。它不仅适用于平稳的环境，而且还适用于非平稳的环境。其估算结果是通过前一次的估计值与新的信息来计算更新的状态值；所以只有先前的估计值需要存储空间。卡尔曼滤波器采用了更高效的线性估计,比过去需要通过计算整个过滤过程中的每一个步骤更有效率。-The well-known Kalman filter, rooted in the state-space linear dynamical systems, an iterative solution of linear recursive optimal filtering problem. It applies not only to the stable environment, but also for non-stationary environment. The estimation result is calculated by the previous estimated value with the new information to update the state value require storage space, so that only the previously estimated values. The Kalman filter uses a more efficient linear estimation, than in the past need to be calculated in the entire filtration process every step more efficient. Platform: |
Size: 67584 |
Author:万达 |
Hits:
Description: the effects of neural
network aided estimation in such receivers are considered. Neural
network acts as a pre-processing block to the estimator.-Orthogonal frequency division multiplexing (OFDM) has
high data rate capacity and lower Inter Symbol Interference (ISI)
and is considered as the best solution for next generation mobile
communication. Multiple Inputs and Multiple Output (MIMO)
antenna system improve reception through spatial diversity and
high end coding. Combining these two, offers high interference
mitigation in wireless receivers. In this paper, the effects of neural
network aided estimation in such receivers are considered. Neural Platform: |
Size: 344064 |
Author:wangxx |
Hits:
Description: This paper presents neural networks based approach
for estimation of the control and operating parameters of Statcom
used for improving voltage profile in a power system, which is
emerging as a major problem in the day-to-day operation of
stressed power systems. Statcom is an important voltage source
converter FACTS device, which can be used in voltage control
mode or reactive power injection mode. For stable operation and
control of power systems it is essential to provide real time
solution to the operator in energy control centers. Artificial neural
networks are proposed here for this task, as they have ability to
synthesize complex mappings accurately and rapidly Platform: |
Size: 506880 |
Author:phdscolar11
|
Hits:
Description: OpenPose人体姿态识别项目是美国卡耐基梅隆大学（CMU）基于卷积神经网络和监督学习并以caffe为框架开发的开源库。可以实现人体动作、面部表情、手指运动等姿态估计。适用于单人和多人，具有极好的鲁棒性。是世界上首个基于深度学习的实时多人二维姿态估计应用，基于它的实例如雨后春笋般涌现。人体姿态估计技术在体育健身、动作采集、3D试衣、舆情监测等领域具有广阔的应用前景，人们更加熟悉的应用就是抖音尬舞机(OpenPost Human Attitude Recognition Project is an open source library developed by Carnegie Mellon University (CMU) based on convolutional neural network and supervised learning and caffe framework. Posture estimation such as human motion, facial expression and finger movement can be realized. It is suitable for single person and multi-person, and has excellent robustness. It is the first real-time multi-person two-dimensional attitude estimation application based on deep learning in the world. Examples based on it have sprung up like mushrooms after a spring rain. Human posture estimation technology has broad application prospects in sports fitness, motion acquisition, 3D fitting, public opinion monitoring and other fields. People are more familiar with the application of tremolo embarrassing dance machine.) Platform: |
Size: 45787136 |
Author:对对对对的 |
Hits:
Search - neural network estimation