Description: 一篇关于SCM的综述性文章，SCM是一种比SVM分类性能更优秀的分类器。以后还会上传几篇有关SCM的文章-on SCM's a review article, SCM than SVM is a better classification performance classifier. After a few will upload the articles SCM Platform: |
Size: 205824 |
Author:ckb |
Hits:
Description: 分别采用感知机算法、最小平方误差算法、线性SVM算法设计分类器，分别画出决策面，并比较性能。-Perceptron algorithm were used, the least square error algorithm, linear SVM classifier algorithm, respectively, making face paint, and compare performance. Platform: |
Size: 77824 |
Author:龚煜 |
Hits:
Description: 基于神经网络的遥感图像分类取得了较好的效果,但存在固有的过学习、易陷入局部极小等缺点.支持向量机机器学习方法,根据结构风险最小化(SRM)原理,表现出很多优于其他传统方法的性能,本研究的基于多类支持向量机分类器的遥感图像分类取得了达95.4 的分类精度.但由于遥感图像分类类别多,所需训练样本较大,人工选择效率较低,为此提出以人工选择初始聚类质心、C均值模糊聚类算法自动标注训练样本的基于多类支持向量机的半监督式遥感图像分类方法,期望能在获得适用的分类精度的基础上有效提高分类效率-Neural net based remote sensing image classification has obtained good results. But neural net has inherent
flaws such as overfitting and local minimums. Support vector machine (SVM), which is based on Structural Risk Min-
imization(SRM), has shown much better performance than most other existing machine learning methods. Using mul-
ti-class SVM classifier high class rate of 95.4 is obtained. But for the class number of remote sensing image is much
great, manually obtaining of training samples is a much time-consuming work. So a multi-class SVM based semi-super-
vised approach is presented. It is choosed that the initial clustering centroids manually first, then label the samples as
the training ones automatically with fuzzy clustering algorithm. It is believed that this method will upgrade the classifi-
cation efficiency greatly with practicable class rate Platform: |
Size: 25600 |
Author:cissy |
Hits:
Description: 对“data1.m”数据，分别采用感知机算法、最小平方误差算法、线性SVM算法设计分类器，分别画出决策面，并比较性能。-The "data1.m" data, respectively, using the perceptron algorithm, the least square error algorithm, the linear SVM algorithm design classifier, respectively, to draw the decision-making surface, and compare performance. Platform: |
Size: 277504 |
Author:刘攀 |
Hits:
Description: 对“data3.m”数据，用其中一半的数据采用非线性SVM算法设计分类器并画出决策面，另一半数据用于测试分类器性能。采用三套核函数，并且比较不同核函数的结果。-To "data3.m" data, which half of the data using nonlinear SVM classification algorithm design and draw the decision-making surface, the other half of the data used to test the classifier performance. Three sets of kernel function, and compare the results of the different kernel functions. Platform: |
Size: 258048 |
Author:刘攀 |
Hits:
Description: SVM神经网络中的参数优化---利用SVM提升分类器性能，很好-Parameter optimization of SVM neural network--- SVM to enhance the performance of the classifier, good Platform: |
Size: 3072 |
Author:suua |
Hits:
Description: 提出了一种基于DCT提取人脸特征技术和支持向量机分类模型的人脸识别方法。利用离
散余弦变换可提取人脸可识别的大部分信息,而支持向量机作为分类器,在处理小样本、高维数等
方面具有独特的优势,且泛化能力很强,无需先验知识。从ORL 人脸库上的实验结果可以看出,
DCT特征提取是很有效的,且SVM的分类性能优于最近邻分类器,同时提高了整个系统的运算速
度。-A face recognition method based on DCT for face feature extraction and support vector machine classification model. Can extract most of the information face recognition using discrete cosine transform and support vector machine as classifier, has unique advantages in dealing with small sample, high dimension and generalization ability, without prior knowledge. As can be seen from the experimental results on the ORL database DCT feature extraction is very effective, and the SVM classification performance better than the nearest neighbor classifier, while increasing the speed of operation of the entire system. Platform: |
Size: 354304 |
Author:罗朝辉 |
Hits:
Description: 《MATLAB神经网络30个案例分析》中的第13个例子，案例13 SVM神经网络中的参数优化---提升分类器性能。希望对大家有一定的帮助！-The MATLAB neural network analysis of 30 cases of example, 13 cases of 13 SVM parameters optimization of neural network classifier performance- ascension. Hope to have certain help to everybody!
Platform: |
Size: 287744 |
Author:杨飞 |
Hits:
Description: SVM神经网络中的参数优化---提升分类器性能,模型精度较高-SVM neural network classifier parameter optimization--- to enhance performance, high accuracy of the model Platform: |
Size: 284672 |
Author:wang yu |
Hits:
Description: 1采用遗传算法对男女生样本数据中的身高，体重，喜欢数学，喜欢文学，喜欢运动，喜欢模式识别共6个特征进行特征选择，并基于所得到的最佳特征采用SVM设计男女生分类器，并计算模型预测性能（包含SE,SP,ACC和AUC ）。提示：可以用6位的0/1进行编码，适应度函数可以考虑类似 。-1 genetic algorithm for boys and girls in the sample data of height, weight, like math, like literature, like sports, like pattern recognition feature a total of six feature selection, design and use of SVM classifier for boys and girls based on the best features of the resulting and calculating prediction performance model (including SE, SP, ACC and AUC). Tip: 6 may be encoded with the 0/1, the fitness function may be considered similar. Platform: |
Size: 10240 |
Author:xiapan |
Hits:
Description: 2采用PCA对男女生样本数据中的身高，体重，喜欢数学，喜欢文学，喜欢运动，喜欢模式识别共6个特征进行特征提取（自己设定选取的特征个数），并基于所得到的特征采用SVM设计男女生分类器，并计算模型预测性能（包含SE,SP,ACC和AUC ）。-2 using PCA for boys and girls in the sample data height, weight, like math, like literature, like sports, like common pattern recognition feature extraction feature 6 (wherein the number of the selected set itself), based on the obtained feature using SVM classifier design boys and girls, and calculating prediction performance model (including SE, SP, ACC and AUC). Platform: |
Size: 12288 |
Author:xiapan |
Hits:
Description: 采用SVM设计男女生分类器。采用的特征包含身高、体重、是否喜欢数学、是否喜欢文学、是否喜欢运动共五个特征。要求：采用平台提供的软件包进行分类器的设计以及测试，尝试不同的核函数设计分类器，采用交叉验证的方式实现对于性能指标的评判（包含SE,SP,ACC和AUC，AUC的计算基于平台的软件包）。-Using SVM classifier is designed for boys and girls. Characterized by the use of include height, weight, whether you like math, whether like literature, whether you like sports a total of five feature. Requirements: The software package provided by the platform classifier design and testing, try different design kernel classifiers, using cross validation manner for performance uation (including SE, SP, ACC and AUC, AUC calculated based on platform software package). Platform: |
Size: 21504 |
Author:xiapan |
Hits:
Description: svm 的参数优化，利用交叉验证法选择最优参数c g，最终提高训练集的分类准确率,更好的提高分类器性能-Svm parameter optimization, the use of cross-validation method to the optimal parameter c g, and ultimately improve the training set classification accuracy，better improve the classifier performance Platform: |
Size: 1024 |
Author:赵珂 |
Hits:
Description: 对SVM神经网络进行参数优化，提升其分类器性能(The parameters of SVM neural network are optimized to improve the performance of classifier) Platform: |
Size: 282624 |
Author:海航
|
Hits:
Description: 采用svm来做分类，一般能得到较满意的结果，但用svm做分类预测时需要调节相关的参数才能得到比较理想的预测分类准确率，那么svm的参数该如何选取？该程序主要说明如何更好地提升分类器性能。(Use svm to do the classification, the general can get more satisfactory results, but when using svm to do classification prediction need to adjust the relevant parameters in order to get the ideal prediction classification accuracy, then svm parameters how to choose? The program mainly shows how to improve the performance of classifier.) Platform: |
Size: 201728 |
Author:wj2511
|
Hits:
Description: 使用SVM分类器来预测乳腺癌病人的预后(特征选择；分类器构建)，评价模型时使用无被交叉验证，性能评价指标包括准确率，AUC，灵敏度，特异度。学会最基本的机器学习方法。可查看分发给大家的代码，以后遇到类似的问题，可用相似的思路和代码。(The SVM classifier was used to predict the prognosis of breast cancer patients (feature selection; classifier construction), and the model was used without cross-validation. Performance evaluation indicators included accuracy, AUC, sensitivity, and specificity. Learn the most basic machine learning methods. You can view the code distributed to everyone, and later encounter similar problems, similar ideas and code can be used.) Platform: |
Size: 1024 |
Author:木葉流光 |
Hits:
Search - svm classifier performance