【数据聚类】改进多元宇宙算法优化密度的噪声应用空间聚类MVO-DBSCAN数据聚类【含Matlab源码 7213期】

Matlab领域

已于 2024-09-04 16:30:39 修改

阅读量263

点赞数 3

文章标签： matlab

于 2024-08-13 16:55:00 首次发布

本文链接：https://blog.csdn.net/weixin_63266434/article/details/141168289

版权

在这里插入图片描述

⛄一、获取代码方式

获取代码方式1：
完整代码已上传我的资源：【数据聚类】基于matlab改进多元宇宙算法优化密度的噪声应用空间聚类MVO-DBSCAN数据聚类【含Matlab源码 7213期】
点击上面蓝色字体，直接付费下载，即可。

获取代码方式2：
付费专栏数据分析（Matlab）

备注：
点击上面蓝色字体付费专栏数据分析（Matlab），扫描上面二维码，付费299.9元订阅海神之光博客付费专栏，凭支付凭证，私信博主，可免费获得5份本博客上传CSDN资源代码（有效期为订阅日起，三天内有效）；
点击CSDN资源下载链接：5份本博客上传CSDN资源代码

⛄二、部分源代码

clc
clear
close all

dataStorage = ‘Datasets’;
resultPath =‘results’;
if ~exist(resultPath,‘dir’)
mkdir(resultPath);
end

DataReposity = {‘artificial’};

Algorithms ={‘MVO-DBSCAN’,‘IMVO1-DBSCAN’,‘IMVO2-DBSCAN’,‘DBSCAN’};
MaxRun = 1;

for rep = 1:length(DataReposity)
rep
dataPath = [dataStorage,DataReposity{rep},‘’];

pathInfo = dir([dataPath,'*.txt']);

for ii = 1:length(pathInfo)
    temp = strsplit(pathInfo(ii).name,'.');
    dataNames{ii} = temp{1};
end

for i = 1:length(dataNames)
    dataName = dataNames{i};
    data = LoadDataset(dataPath,DataReposity{rep},dataName);
    data.dataName = dataName;
    
    for j = 1:length(Algorithms)
        AlgName = Algorithms{j}
        inits  = initialization(DataReposity{rep},data,AlgName);
        
        for run = 1:MaxRun
            
            if strcmp(AlgName,'MVO-DBSCAN')
                
                % execute clustering method on the problem
                out = MVO_DBSCAN(inits.prob,inits.params);
                clus = out.ptsC;
                C = out.C;
                Nc = out.Nc;
                
            end
            
            if strcmp(AlgName,'IMVO1-DBSCAN')
                
                % execute clustering method on the problem
                out = IMVO1_DBSCAN(inits.prob,inits.params);
                clus = out.ptsC;
                C = out.C;
                Nc = out.Nc;
                
            end
            
            if strcmp(AlgName,'IMVO2-DBSCAN')
                
                % execute clustering method on the problem
                out = IMVO2_DBSCAN(inits.prob,inits.params);
                clus = out.ptsC;
                C = out.C;
                Nc = out.Nc;
                
            end
            
            if strcmp(AlgName,'DBSCAN')
                
                % execute clustering method on the problem
                X = data.X; % points
                [ptsC,C,Nc,Pvisit] = dbscan(X', inits.E, inits.minPts);
                clus = ptsC;
                
            end
            
            %% save results
            savePath = [resultPath,DataReposity{rep},'\',AlgName,'\'];
            if ~exist(savePath,'dir')
                mkdir(savePath);
            end
            
            if run == 1
                if  size(data.X,2) == 2
                    h = PlotClusteringResults(data.X,clus);
                    print(h,[savePath,dataName],'-djpeg','-r300')
                    close(h);
                end
                if strcmp(AlgName,'DBSCAN')
                    save([savePath,dataName,'_outs'],'clus','C','Nc');
                else
                    GlobalBest = out.GlobalBest;
                    save([savePath,dataName,'_outs'],'clus','C','Nc','GlobalBest');
                end
                
            end
            
            %%  Evaluation
            [results{i}.MeanAccuracy(run,j),results{i}.err(run,j)] = ...
                ClusteringEvaluation1(clus,data.y);
            [results{i}.F_measure(run,j)] =  ClusteringEvaluation2(clus,data.y);
            
        end
    end
    results{i}.dataName = dataName;
    
end
save([resultPath,DataReposity{rep},'\','results'],'results');
% bar plot
resultName = 'results';
barPlots([resultPath,DataReposity{rep},'\'],resultName,dataNames,Algorithms);
clear results dataNames

end