classify a instance(supervised learning)

Help
2012-07-07
2012-09-14
  • hi

    i want to classify a instance of a dataset,the class of my instances is
    categorical,how can with predict return a categorical?

    but my program do not work true

    please help me

        GMatrix* pData= GMatrix::loadArff("/home/t3/yousefian/dataset/darpa99/Extract-features/W4-2.arff"); // load some data
        //Holder hm(pM); // make sure it will be cleaned up later
        Holder<GMatrix> hData(pData);
        cout<<"pdata"<<"RowSize: "<<pData->rows()<<"ColSize: "<<pData->cols()<<endl;//return: pdataRowSize: 19158 ColSize: 20
        GMatrix* pFeatures = pData->cloneSub(0, 0, pData->rows(), pData->cols() - 1);
        cout<<"pFeatures"<<"RowSize: "<<pFeatures->rows()<<"ColSize: "<<pFeatures->cols()<<endl;//return: pFeaturesRowSize: 19158 ColSize: 19
            Holder<GMatrix> hFeatures(pFeatures);
            GMatrix* pLabels = pData->cloneSub(0, pData->cols() - 1, pData->rows(), 1);
            cout<<"pLables"<<"RowSize: "<<pLabels->rows()<<"ColSize: "<<pLabels->cols()<<endl;//return: pLablesRowSize: 19158 ColSize: 1
            Holder<GMatrix> hLabels(pLabels);
    
            GRand rand(0);
            GKNN modell(rand);
            modell.setNeighborCount(3);
            modell.train(*pFeatures, *pLabels);
    
            double pOut[1];
    
                modell.predict(pFeatures->row(37), pOut);
    

    thanks

     
  • Mike Gashler
    Mike Gashler
    2012-07-20

    Your code looks right to me. Can you describe how your actual results differ
    from your expected results?

     
  • thanks for you attention

    how can return lable of row(37)??

    cout<< pOut dont worked!!

     


Anonymous


Cancel   Add attachments