Location:
Search - 分类精度
Search list
Description: 有向无环图支持向量(DAG-SVMS)多类分类方法,是一种新的多类分类方法。该方法采用了最小超球体类包含作为层次分类依据。试验结果表明,采用该方法进行多类分类,跟已有的分类方法相比有更高的分类精度。
Platform: |
Size: 5049 |
Author: 苏苏 |
Hits:
Description: 基于weka平台的数据挖掘方法，改进特征基因选择过程，使得癌症分类精度更高
Platform: |
Size: 3584568 |
Author: 王莲花 |
Hits:
Description: 针对遥感影像的光谱信息进行分类，并评价分类精度。但需要人为调整数组的大小，来控制输入变量，将训练样本和检验样本存为.txt格式的文件，执行即可得到分类后属于各个类别的概率，以及混淆矩阵。
Platform: |
Size: 25531 |
Author: Una |
Hits:
Description: 影像分类精度评价算法——混淆矩阵，评价漏分和错分情况
Platform: |
Size: 28177 |
Author: 海燕 |
Hits:
Description: BP学习算法应用——模式分类
应用动量BP学习算法对UCI提供的经典数据库——鸢尾属植物数据库进行分类，速度快，精度高。iris.arff为数据库文件，可用Weka数据挖掘软件打开。Iris.csv为源代码读取的数据文件，通过Weka软件转换得到。
将源文件Iris_classify.m和Iris.csv文件放入matlab的work文件夹中直接运行即可。-BP learning algorithm applications- the application of pattern classification momentum BP learning algorithm to provide a classic UCI database- iris database classification, high speed and accuracy. iris.arff for the database file that can be used to open Weka data mining software. Iris.csv for the source code to read the data files have been converted through Weka software. Iris_classify.m the source document and document Iris.csv Add matlab folder of work can be directly run.
Platform: |
Size: 5120 |
Author: Michael_M |
Hits:
Description: 有向无环图支持向量(DAG-SVMS)多类分类方法,是一种新的多类分类方法。该方法采用了最小超球体类包含作为层次分类依据。试验结果表明,采用该方法进行多类分类,跟已有的分类方法相比有更高的分类精度。
-Directed acyclic graph support vector (DAG-SVMS) multi-category classification methods, is a new multi-category classification methods. The method uses the smallest category of super-sphere that contains the level of classification as a basis. The experimental results show that using the method of multiclass classification with the classification method has been compared to a higher classification accuracy.
Platform: |
Size: 5120 |
Author: 苏苏 |
Hits:
Description: 提出了一种通过遗传算法(GA)对单个分类器进行优化以及对多个分类器进行组合优化的方法.该方法使用叠加(stacking)的策略.经典的叠加策略分为两步,该方法将遗传算法作为叠加策略的第2步.实验结果表明,遗传算法可以较好地完成优化任务,同单个分类器比较,它可以提高分类的精度.在对分类器进行组合优化方面,它得到比单个分类器更高的精度以及使分类结果具有更好的可理解性.
关 键 词: 分类 遗传算法 优化 机器学习 数据挖掘 分类规则.-err
Platform: |
Size: 68608 |
Author: limingxian |
Hits:
Description: 利用主成分分析法对BP神经网络的输入参数进行降维，然后进行网络的训练，PCA-BP处理的结果同单一的bp相比，不仅提高了网络的收敛速度，而且提高了网络对预测数据分类的精度-Using principal component analysis method of BP neural network for dimensionality reduction of input parameters, and then training the network, PCA-BP deal with the results of a single bp, compared with not only improve the network convergence rate, and improve the network prediction data Classification accuracy
Platform: |
Size: 1024 |
Author: 娜娜 |
Hits:
Description: 基于weka平台的数据挖掘方法，改进特征基因选择过程，使得癌症分类精度更高-Weka platform based on data mining methods, to improve the characteristics of gene selection process, making cancer a higher classification accuracy
Platform: |
Size: 3584000 |
Author: 王莲花 |
Hits:
Description: 针对遥感影像的光谱信息进行分类，并评价分类精度。但需要人为调整数组的大小，来控制输入变量，将训练样本和检验样本存为.txt格式的文件，执行即可得到分类后属于各个类别的概率，以及混淆矩阵。-Spectral imaging for remote sensing information classification, and to evaluate the classification accuracy. However, the need to artificially adjust the size of the array to control input variables, the training samples and the samples tested for the depositors. Txt files, the implementation can be sorted and the probability of belonging to each category, as well as the confusion matrix.
Platform: |
Size: 347136 |
Author: Una |
Hits:
Description: 影像分类精度评价算法——混淆矩阵，评价漏分和错分情况-Image Classification Accuracy Assessment Algorithm- confusion matrix, evaluation of leakage points and the wrong sub-situation
Platform: |
Size: 372736 |
Author: 海燕 |
Hits:
Description: 很好用的svm工具箱，dagsvm对于图像分类效果非常好，分类精度很高，代码清洗简单。-Good use of SVM Toolbox, dagsvm for image classification effect is very good, very high classification accuracy, code cleaning easy.
Platform: |
Size: 5120 |
Author: tang |
Hits:
Description: 分类判别中，bayes判别的确具有明显的优势，与模糊，灰色，物元可拓相比，判别准确率一般都会高些，而BP神经网络由于调试麻烦，在调试过程中需要人工参与，而且存在明显的问题，局部极小点和精度与速度的矛盾，以及训练精度和仿真精度间的矛盾，等，尽管是非线性问题的一种重要方法，但是在我们项目中使用存在一定的局限，基于此，最近两天认真的研究了bayes判别，并写出bayes判别的matlab程序，与spss非逐步判别计算结果一致。-Classified Identifying, bayes discriminant does have a distinct advantage, with the fuzzy, gray, matter-element and extension compared to determine the exact rate will be higher in general, and the BP neural network trouble as a result of debugging, in the need to manually debug the process of participation, but also obvious problems, the local minimum point and the accuracy and speed of contradictions, as well as simulation training accuracy and precision of the conflict between, and so on, in spite of nonlinear problems is an important method, but the use of our project there are certain limitations, based on the Here, seriously the last couple of days to study the discriminant bayes and bayes discriminant of matlab to write procedures, and non-spss stepwise discriminant calculation results.
Platform: |
Size: 4096 |
Author: lili |
Hits:
Description: 大名鼎鼎的方帅的博士学位论文---目前,计算机智能视频监控在理论和应用上都面临着很多难题,国内外大批学者投身于该领域的研究和探索,并且取得了大量的成果.本文是在这些成果的基础上,对计算机智能视频监控系统的关键技术进行研究.主要贡献可概括如下:首先,对目标检测技术进行了研究,并提出了一种基于背景建模的运动目标检测算法.利用统计的方法建立了基于颜色和颜色梯度的背景模型,并实时地对背景模型进行更新,最后将这两种背景模型综合考虑对目标进行了有效的检测.接着,研究了复杂背景下多目标跟踪问题,提出了基于蒙特卡罗粒子滤波器的复杂背景下多目标跟踪算法.然后,提出了目标分类算法.用有向无环图的多类支持向量机对目标进行分类,该分类器使用少量的样本,就可以得到较好的分类精度.最后,针对视频监控中的行为理解进行了探讨性的研究.建立一个行为理解的系统,这个系统由场景模型、行为模型和行为自动识别三部分组成.
-At present, the computer intelligence video frequency monitoring is facing many difficult problems in the theory and the application, the domestic and foreign large quantities of scholars join in this domain research and the exploration, and has made the massive progresses. This article is in these achievement foundation, conducts the research to the computer intelligence video frequency supervisory system s key technologies. The main contribution may summarize as follows: First, has conducted the research to the object detection technology, and proposed one kind based on the background modelling movement object detection algorithm. Has established using the statistical method based on the color and the color gradient background model, and real-time carries on the renewal to the background model, finally has carried on these two kind of background model overall evaluation to the goal the effective examination. Then, has studied under the complex background the multi-objective track que
Platform: |
Size: 5263360 |
Author: 王一 |
Hits:
Description: 编写朴素贝叶斯分类器，对测试集进行分类预测，并计算分类精度(The naive Bias classifier is compiled, and the test set is classified and predicted, and the classification accuracy is calculated)
Platform: |
Size: 1024 |
Author: 王FANFANFAN |
Hits:
Description: 一般的支持向量机只支持二分类，使用libsvm可以实现多分类，原理也是基于二分类，然后在使用投票机制，经测验，libsvm的分类精度可达85%以上(Multi class supported by libsvm,after testing, the classification accuracy can reach 85%.)
Platform: |
Size: 466944 |
Author: 韩alan |
Hits:
Description: 二元粒子群优化（BPSO）用于特征选择任务，可以选择潜在特征，提高分类精度。(binary particle swarm optimization (BPSO) for feature selection tasks, which can select the potential features to improve the classification accuracy.)
Platform: |
Size: 61440 |
Author: 皮张 |
Hits:
Description: 通过已有的混淆矩阵计算总体分类精度、期望精度以及Kappa系数
Platform: |
Size: 234 |
Author: 1135063213@qq.com |
Hits:
Description: 基于SVM的疲劳驾驶系统。基于神经网络的非接触式疲劳驾驶检测已成为当前针对疲劳驾驶检测领域炙手可热的研究方向。它有效解决了接触式疲劳检测方法给驾驶员带来的干扰以及单一信号源对于反映疲劳程度可靠性低的问题，同时通过设计神经网络模型对多源信息进行分类，实现对疲劳状态的高精度和高速度的检测。选取合适的特征值对网络检测准确率以及准确反映疲劳程度至关重要。基于驾驶员生理信号检测可靠性和准确性较高。(Fatigue driving system based on SVM)
Platform: |
Size: 89088 |
Author: pakchoi007 |
Hits:
Description: 基于python，利用主成分分析（PCA）和K近邻算法（KNN）在MNIST手写数据集上进行了分类。
经过PCA降维，最终的KNN在100维的特征空间实现了超过97%的分类精度。(Based on python, it uses principal component analysis (PCA) and K nearest neighbor algorithm (KNN) to classify on the MNIST handwritten data set.
After PCA dimensionality reduction, the final KNN achieved a classification accuracy of over 97% in a 100-dimensional feature space.)
Platform: |
Size: 11599872 |
Author: 曲小刀 |
Hits: