Description: 采用BP神经网络算法实现的大写金融汉字识别软件的核心识别功能，适合学习。-BP neural network algorithm to achieve the financial capital of the Chinese character recognition software to identify the core features for learning. Platform: |
Size: 54398 |
Author:sinan |
Hits:
Description: 一种基于蚁群聚类的径向基神经网络
提出了一种基于蚁群聚类算法的径向基神经网络. 利用蚁群算法的并行寻优特征和挥发系
数方法的自适应更改信息量的能力,并以球面聚类的方式确定了径向基神经网络中基函数的位置,
同时通过比较隐层神经元的相似性、合并相似性较为接近的2 个神经元来约简隐含层的神经元,以
达到简化径向基神经网络结构的目的. 实验比较了几种不同聚类算法的径向基神经网络,结果表
明,所提神经网络的整体训练时间至少可缩短40 ％ ,学习的准确率可提高1 ％以上,而且网络结构
更加精简.-An Ant clustering RBFNN presents a clustering algorithm based on the ant colony RBFN Liaison. Ant use the parallel algorithm optimization features and volatile coefficient method of adaptive changes in the volume of information, spherical cluster and the method of RBF neural network-based functions of the position, By comparing the same time hidden layer neurons of similarity, Merger similarity is closer to the two neurons to Jane about the hidden layer neurons, to achieve simplification RBF neural network structure purposes. Experimental comparison of several different clustering algorithm RBF neural network, The results show that the neural network's overall training time can be shortened at least 40%. Learning the accuracy rate can be increased by 1%, and the network is more streaml Platform: |
Size: 292558 |
Author:zhxj |
Hits:
Description: 神经网络（模式识别模块）软件开发包!!!可以识别任何事物，适用于图像识别、语音识别、自动控制等诸多领域，具有简单、易学、开发周期短、识别率高等特点。-neural network (pattern recognition module) software development kit! ! ! Nothing can be identified and applied to image recognition, voice recognition, automatic control and many other areas, which is simple, easy to learn, short development cycle, features high recognition rate. Platform: |
Size: 139398 |
Author:於俊 |
Hits:
Description: 用三层小波神经网络实现的与文本无关说话人识别。（识别部分）。输入的是语音特征，输出的是识别结果。训练用的语音特征要事先提取出。-with three wavelet neural network has nothing to do with the text of Speaker Recognition. (Recognition). The admission of voice features, the output is the result of recognition. Training voice features prior to extract. Platform: |
Size: 6193 |
Author:白莹 |
Hits:
Description: 人工神经网络(ANN)的泛化特性是神经网络最重要的特性,同时也是最不容易保证的特性。本程序对改进泛化的神经网络算法以及新兴的机器学习算法——支持向量机算法进行研究,-Artificial Neural Network (ANN) the generalization characteristics of neural networks are the most important characteristics, but also not easy to guarantee the most features. This procedure for improving the generalization of neural network algorithm, as well as the emerging machine learning algorithms- Support Vector Machine algorithm research, Platform: |
Size: 7168 |
Author:王旭 |
Hits:
Description: 利用BP神经网络对遥感图像进行地物的自动识别-The use of BP neural network of remote sensing images of the automatic recognition of features Platform: |
Size: 683008 |
Author:harvey |
Hits:
Description: 基于BP神经网络的遥感图像分类代码。从样本中提取崇明岛东滩十种地物的光谱特征，并训练BP网络，再利用网络进行分类-BP neural network-based remote sensing image classification code. Extracted from samples of 10 kinds of Chongming Island, Dongtan features of the spectral characteristics and to train BP network, and then use the network to classify Platform: |
Size: 1413120 |
Author:李琛 |
Hits:
Description: 利用bp神经网络对遥感图像进行分类，输入样本值后，根据样本值对遥感图像不同的地物进行分类，分类后计算每种地物所占面积-The use of bp neural network classification of remote sensing images, enter a sample value, based on the sample value of the different features of remote sensing image classification, classification features calculated for each share of area Platform: |
Size: 3072 |
Author:sun |
Hits:
Description: LmNet PF 神经网络预测平台是公司基于最优神经网络算法(Levenberg-Marquardt动量项法)开发的通用预测平台工具。它是针对用户进行预测需要，快速构建神经网络应用的通用预测平台，它能解决包括销售量预测、销售价格预测、成本预测、市场潜力预测、新产品价格预测等方面的预测分析。功能包括：新建、修改网络模型；网络训练；网络仿真；误差分析；专家样本数据自动生成；节点配置；数据归一化处理；网络参数初始化设置等。~..~
-Neural Network Prediction LmNet PF platform is based on the optimal neural network algorithm (Levenberg-Marquardt momentum of Law) predict the development of common tools platform. It is for users to predict the need for rapid construction of the universal application of neural network forecasting platform, it can solve, including sales forecasts, sales price forecasts, cost projections, the forecast market potential of new products, price forecasts for the areas of predictive analysis. Features include: new, modify the network model network training network simulation error analysis experts automatically generate sample data node configuration data normalization network settings such as initialization parameters. ~ .. ~ Platform: |
Size: 14720000 |
Author:fbia |
Hits:
Description: Based on the information processing functionalities of spiking neurons, a spiking neural network model is proposed to extract features from a visual image. The network is constructed with a conductance-based integrate-and-fire neuron model and a set of specific receptive fields. The properties of the network are detailed in this paper. Simulation results show that the network is able to perform image feature extraction within a time interval of 100 ms. This processing time is consistent with the human visual system. The demonstrations show how the network can extract right-angle contours in a visual image. Based on this principle, many other image features can be extracted by analogy. The parallel processing mechanism of this network model is very promising for a hardware implementation based on VLSI or FPGA technology. Platform: |
Size: 288768 |
Author:sacoura31 |
Hits:
Description: 神经网络做的图像模式识别，用的是神经网络分割特征，自己研究的，比较有新意-Neural network to do the image pattern recognition, using a neural network segmentation features, its own research, more new ideas Platform: |
Size: 124928 |
Author:dingzi |
Hits:
Description: 火灾视觉特征的提取是视觉火灾探测中的关键问题. 我们主要研究色彩、纹理以及轮廓脉动
等特征的提取,并提出一种度量轮廓脉动信息的距离模型,该模型在规格化的傅立叶描述子空间能
够准确地度量这种时空闪烁特征. 实验结果表明,该方法具有比较好的鲁棒性,有助于提高视觉火
灾探测的准确率、降低误报漏报率.-Based on investigating color , text ure and temporal feat ures for vision based fire detection , a
distance model of contour fluct uation between two successive f rames in t he normalized Fourier descriptor s
domain was presented to measure t his time varying contour fluct uation feat ure of flame. The model of
contour fluct uation is effective and robust for fire recognition. To f urt her reduce fal se alarms , several
features ext racted according to color , text ure and the distance model were toget her regarded as a joint
feature vector for artificial neural network to detect fire. Experiment s show t hat the algorithm is effective
and robust , and t hat it is significant for improving accuracy and reducing fal se alarms. Platform: |
Size: 819200 |
Author:陈卿 |
Hits:
Description: High information redundancy and correlation in face images result in efficiencies when such images are used directly for recognition. In this paper, discrete cosine transforms are used to reduce image information redundancy because only a subset of the transform coefficients are necessary to preserve the most important facial features such as hair outline, eyes and mouth. We demonstrate experimentally that when DCT coefficients are fed into a backpropagation neural network for classification, a high recognition rate can be achieved by using a very small proportion of transform coefficients. This makes DCT-based face recognition much faster than other approaches.-High information redundancy and correlation in face images result in inefficiencies when such images are used directly for recognition. In this paper, discrete cosine transforms are used to reduce image information redundancy because only a subset of the transform coefficients are necessary to preserve the most important facial features such as hair outline, eyes and mouth. We demonstrate experimentally that when DCT coefficients are fed into a backpropagation neural network for classification, a high recognition rate can be achieved by using a very small proportion of transform coefficients. This makes DCT-based face recognition much faster than other approaches. Platform: |
Size: 25600 |
Author:mhm |
Hits:
Description: 该代码的功能主要是识别数字。采用的算法是建立人工神经网络的模型，通过对样本的训练而保存全权值（及数字的特征），训练结束后能识别图片上的数字（256位色），相信对于研究神经网络模型的很有帮助，同时该代码还涉及到图像处理部分的内容，比如二值化，梯度锐化，去噪声等算法。-The main function is to identify the code number. The algorithm used is based on artificial neural network model, through training and save the full sample values (and the number of features), training to recognize the picture after the number (256-bit color), I believe that neural network model for the study was helpful, while the code is also involved in image processing part of the contents, such as binary, gradient sharpening algorithms to noise. Platform: |
Size: 3229696 |
Author:沈晓伟 |
Hits:
Description: Extension Neural Networks by Wang for character recognition using the given features of any particular attribute of a genetic entity Platform: |
Size: 211968 |
Author:Surbhi |
Hits:
Description: Key Features
* Neural network design, training, and simulation
* Pattern recognition, clustering, and data-fitting tools
* Supervised networks including feedforward, radial basis, LVQ, time delay, nonlinear autoregressive (NARX), and layer-recurrent
* Unsupervised networks including self-organizing maps and competitive layers
* Preprocessing and postprocessing for improving the efficiency of network training and assessing network performance
* Modular network representation for managing and visualizing networks of arbitrary size
* Routines for improving generalization to prevent overfitting
* Simulink® blocks for building and evaluating neural networks, and advanced blocks for control systems applications-Key Features
* Neural network design, training, and simulation
* Pattern recognition, clustering, and data-fitting tools
* Supervised networks including feedforward, radial basis, LVQ, time delay, nonlinear autoregressive (NARX), and layer-recurrent
* Unsupervised networks including self-organizing maps and competitive layers
* Preprocessing and postprocessing for improving the efficiency of network training and assessing network performance
* Modular network representation for managing and visualizing networks of arbitrary size
* Routines for improving generalization to prevent overfitting
* Simulink® blocks for building and evaluating neural networks, and advanced blocks for control systems applications Platform: |
Size: 252928 |
Author:bacha |
Hits:
Description: 摘要:回顾了神经网络理论发展的历史和现状。在此基础上，介绍并讨论了20世纪如年代神经网络研究的一些新进展。根据神经网络研究的特点，对人工神经网络今后的发展前景作了一定的评述，
人工神经网络的研究与发展必将对现代科学技术产生深远的影响.配准中基于像素和基于特征两类主要方法的优缺点 深入研究了互信息理论在医学图像配准领域的应用和实现机理以及互信息作为相似性测度的优缺点。本文着
重深入研究了Renyi广义互信息及其在图像配准中的作用，并针对互信息计算的复杂性，引入了广义近邻图嫡估计理论，并成功应用于多模医学图像配准中。
-Abstract: The review of the history and current status of the development of neural network theory. On this basis, some new developments in the 20th century, such as the age of neural network research is presented and discussed. According to the characteristics of neural network research, made some comment on the future prospects for the development of artificial neural network, artificial neural network research and development is bound to have a profound impact on modern science and technology. Registration based on the pixel-based features of the two main types The advantages and disadvantages of the method -depth study of the mutual information theory in the field of medical image registration and achieve mechanism and the advantages and disadvantages of mutual information as a similarity measure. The article focuses on in-depth study the Renyi generalized mutual information and its role in image registration, and the introduction of mutual information calculation complexity, the t Platform: |
Size: 45056 |
Author:朙朙 |
Hits:
Description: Major features
BNT supports many types of conditional probability distributions (nodes), and it is easy to add more.
Tabular (multinomial)
Gaussian
Softmax (logistic/ sigmoid)
Multi-layer perceptron (neural network)
Noisy-or
Deterministic
Platform: |
Size: 12275013 |
Author:SunStacy |
Hits:
Description: 深度学习python实现，并附有MNIST上的测试程序，准确率98 以上-Deep learning learns low and high-level features large amounts of unlabeled data, improving classification on different, labeled, datasets. Deep learning can achieve an accuracy of 98 on the MNIST dataset.
Platform: |
Size: 2017280 |
Author:孙立立 |
Hits:
Search - features neural network