5-進一步研究使用libsvm的筆記

研究libsvm 直到能夠將其 納爲己用

大部分代碼來自libsvm壓縮包裏的文件，通過分析裏面的代碼得出其用法。

Title 1—svm_parameter參數的設置：

svm_parameter 該類用來設置libsvm的一些配置參數，參數有點多，具體的參數說明如下:

private static void exit_with_help()
    {
        System.out.print(
         "Usage: svm_train [options] training_set_file [model_file]\n"
        +"options:\n"
        +"-s svm_type : set type of SVM (default 0)\n"
        +"  0 -- C-SVC      (multi-class classification)\n"
        +"  1 -- nu-SVC     (multi-class classification)\n"
        +"  2 -- one-class SVM\n"
        +"  3 -- epsilon-SVR    (regression)\n"
        +"  4 -- nu-SVR     (regression)\n"
        +"-t kernel_type : set type of kernel function (default 2)\n"
        +"  0 -- linear: u'*v\n"
        +"  1 -- polynomial: (gamma*u'*v + coef0)^degree\n"
        +"  2 -- radial basis function: exp(-gamma*|u-v|^2)\n"
        +"  3 -- sigmoid: tanh(gamma*u'*v + coef0)\n"
        +"  4 -- precomputed kernel (kernel values in training_set_file)\n"
        +"-d degree : set degree in kernel function (default 3)\n"
        +"-g gamma : set gamma in kernel function (default 1/num_features)\n"
        +"-r coef0 : set coef0 in kernel function (default 0)\n"
        +"-c cost : set the parameter C of C-SVC, epsilon-SVR, and nu-SVR (default 1)\n"
        +"-n nu : set the parameter nu of nu-SVC, one-class SVM, and nu-SVR (default 0.5)\n"
        +"-p epsilon : set the epsilon in loss function of epsilon-SVR (default 0.1)\n"
        +"-m cachesize : set cache memory size in MB (default 100)\n"
        +"-e epsilon : set tolerance of termination criterion (default 0.001)\n"
        +"-h shrinking : whether to use the shrinking heuristics, 0 or 1 (default 1)\n"
        +"-b probability_estimates : whether to train a SVC or SVR model for probability estimates, 0 or 1 (default 0)\n"
        +"-wi weight : set the parameter C of class i to weight*C, for C-SVC (default 1)\n"
        +"-v n : n-fold cross validation mode\n"
        +"-q : quiet mode (no outputs)\n"
        );
        System.exit(1);
    }

也可以使用默認的參數，默認的參數爲: 需要注意的是爲gamma參數爲1/num_features而不是0

// default values
        param.svm_type = svm_parameter.C_SVC;
        param.kernel_type = svm_parameter.RBF;
        param.degree = 3;
        param.gamma = 0;    // 1/num_features
        param.coef0 = 0;
        param.nu = 0.5;
        param.cache_size = 100;
        param.C = 1;
        param.eps = 1e-3;
        param.p = 0.1;
        param.shrinking = 1;
        param.probability = 0;
        param.nr_weight = 0;
        param.weight_label = new int[0];
        param.weight = new double[0];
        cross_validation = 0;

完整參考代碼如下 ：

private void parse_command_line(String argv[])
    {
        int i;
        svm_print_interface print_func = null;  // default printing to stdout
        param = new svm_parameter();
        // default values
        param.svm_type = svm_parameter.C_SVC;
        param.kernel_type = svm_parameter.RBF;
        param.degree = 3;
        param.gamma = 0;    // 1/num_features
        param.coef0 = 0;
        param.nu = 0.5;
        param.cache_size = 100;
        param.C = 1;
        param.eps = 1e-3;
        param.p = 0.1;
        param.shrinking = 1;
        param.probability = 0;
        param.nr_weight = 0;
        param.weight_label = new int[0];
        param.weight = new double[0];
        cross_validation = 0;
        // parse options
        for(i=0;i<argv.length;i++)
        {
            if(argv[i].charAt(0) != '-') break;
            if(++i>=argv.length)
                exit_with_help();
            switch(argv[i-1].charAt(1))
            {
                case 's':
                    param.svm_type = atoi(argv[i]);
                    break;
                case 't':
                    param.kernel_type = atoi(argv[i]);
                    break;
                case 'd':
                    param.degree = atoi(argv[i]);
                    break;
                case 'g':
                    param.gamma = atof(argv[i]);
                    break;
                case 'r':
                    param.coef0 = atof(argv[i]);
                    break;
                case 'n':
                    param.nu = atof(argv[i]);
                    break;
                case 'm':
                    param.cache_size = atof(argv[i]);
                    break;
                case 'c':
                    param.C = atof(argv[i]);
                    break;
                case 'e':
                    param.eps = atof(argv[i]);
                    break;
                case 'p':
                    param.p = atof(argv[i]);
                    break;
                case 'h':
                    param.shrinking = atoi(argv[i]);
                    break;
                case 'b':
                    param.probability = atoi(argv[i]);
                    break;
                case 'q':
                    print_func = svm_print_null;
                    i--;
                    break;
                case 'v':
                    cross_validation = 1;
                    nr_fold = atoi(argv[i]);
                    if(nr_fold < 2)
                    {
                        System.err.print("n-fold cross validation: n must >= 2\n");
                        exit_with_help();
                    }
                    break;
                case 'w':
                    ++param.nr_weight;
                    {
                        int[] old = param.weight_label;
                        param.weight_label = new int[param.nr_weight];
                        System.arraycopy(old,0,param.weight_label,0,param.nr_weight-1);
                    }
                    {
                        double[] old = param.weight;
                        param.weight = new double[param.nr_weight];
                        System.arraycopy(old,0,param.weight,0,param.nr_weight-1);
                    }
                    param.weight_label[param.nr_weight-1] = atoi(argv[i-1].substring(2));
                    param.weight[param.nr_weight-1] = atof(argv[i]);
                    break;
                default:
                    System.err.print("Unknown option: " + argv[i-1] + "\n");
                    exit_with_help();
            }
        }
        svm.svm_set_print_string_function(print_func);
        // determine filenames
        if(i>=argv.length)
            exit_with_help();
        input_file_name = argv[i];
        if(i<argv.length-1)
            model_file_name = argv[i+1];
        else
        {
            int p = argv[i].lastIndexOf('/');
            ++p;    // whew...
            model_file_name = argv[i].substring(p)+".model";
        }
    }

   Title 2 : svm_problem 用來保存樣本實例的

             prob.l l用來保存樣本的數量

             prob.x x是一個svm_node類型的二維數組，每一行代表一個樣本，svm_node實例裏的value代表屬性值，index代表屬性的序號。

             prob.y y是一個double類型的一維數組，用來保存樣本的類別標籤。

             prob.y[i] 與 prob.x[i][]構成了一個完整的樣本，一個保存類別標籤，一個保存屬性值。

private void read_problem() throws IOException
    {
        BufferedReader fp = new BufferedReader(new FileReader(input_file_name));
        Vector<Double> vy = new Vector<Double>();
        Vector<svm_node[]> vx = new Vector<svm_node[]>();
        int max_index = 0;
        while(true)
        {
            String line = fp.readLine();
            if(line == null) break;
            StringTokenizer st = new StringTokenizer(line," \t\n\r\f:");
            vy.addElement(atof(st.nextToken()));
            int m = st.countTokens()/2;
            svm_node[] x = new svm_node[m];
            for(int j=0;j<m;j++)
            {
                x[j] = new svm_node();
                x[j].index = atoi(st.nextToken());
                x[j].value = atof(st.nextToken());
            }
            if(m>0) max_index = Math.max(max_index, x[m-1].index);
            vx.addElement(x);
        }
        prob = new svm_problem();
        prob.l = vy.size();
        prob.x = new svm_node[prob.l][];
        for(int i=0;i<prob.l;i++)
            prob.x[i] = vx.elementAt(i);
        prob.y = new double[prob.l];
        for(int i=0;i<prob.l;i++)
            prob.y[i] = vy.elementAt(i);
        if(param.gamma == 0 && max_index > 0)
            param.gamma = 1.0/max_index;
        if(param.kernel_type == svm_parameter.PRECOMPUTED)
            for(int i=0;i<prob.l;i++)
            {
                if (prob.x[i][0].index != 0)
                {
                    System.err.print("Wrong kernel matrix: first column must be 0:sample_serial_number\n");
                    System.exit(1);
                }
                if ((int)prob.x[i][0].value <= 0 || (int)prob.x[i][0].value > max_index)
                {
                    System.err.print("Wrong input format: sample_serial_number out of range\n");
                    System.exit(1);
                }
            }
        fp.close();
    }

一般的分類步驟如下:

利用訓練樣本生成一個分類模型，然後利用該分類模型對待分類的樣本進行分類

我需要的libsvm的使用步驟：

step 1:設置svm_parameter (注，gamma參數 的使用是在 readproblem函數裏面使用的以及當kernel_type == PRECOMPUTED的時候也是在readproblem裏面生效的,etc...)

step 2:將自己的樣本存儲到svm_problem對象中

step 3:利用svm_problem和svm_paramter通過svm.svm_train函數可以訓練出一個svm_model

       public static svm_model svm_train(svm_problem prob, svm_parameter param)

       svm.java裏面的參考代碼 svm_model submodel = svm_train(subprob,param);  

step 4:有了模型之後我們就可以進行分類了:

      根據參數的不同調用的分類函數有所不同，但是主要有兩個成員函數可以使用 svm_predict_probalility以及svm_predict 返回值就是分類的結果

       public static double svm_predict_values(svm_model model, svm_node[] x, double[] dec_values)

       public static double svm_predict(svm_model model, svm_node[] x)

      參考代碼如下

if(param.probability==1 &&
               (param.svm_type == svm_parameter.C_SVC ||
                param.svm_type == svm_parameter.NU_SVC))
            {
                double[] prob_estimates= new double[svm_get_nr_class(submodel)];
                for(j=begin;j<end;j++)
                    target[perm[j]] = svm_predict_probability(submodel,prob.x[perm[j]],prob_estimates);
            }
            else
                for(j=begin;j<end;j++)
                    target[perm[j]] = svm_predict(submodel,prob.x[perm[j]]);

svm.java 源碼裏的開放接口 從字面意義上可以看出點門路...應該有相關文檔吧 沒去找哈- -以後有需要再去研究

public static svm_model svm_train(svm_problem prob, svm_parameter param)

public static void svm_cross_validation(svm_problem prob, svm_parameter param, int nr_fold, double[] target)

public static int svm_get_svm_type(svm_model model)

public static int svm_get_nr_class(svm_model model)

public static void svm_get_labels(svm_model model, int[] label)

public static void svm_get_sv_indices(svm_model model, int[] indices)

public static int svm_get_nr_sv(svm_model model)

public static double svm_get_svr_probability(svm_model model)

public static double svm_predict_values(svm_model model, svm_node[] x, double[] dec_values)

public static double svm_predict(svm_model model, svm_node[] x)

public static double svm_predict_probability(svm_model model, svm_node[] x, double[] prob_estimates)

public static void svm_save_model(String model_file_name, svm_model model) throws IOException

public static svm_model svm_load_model(String model_file_name) throws IOException

public static svm_model svm_load_model(BufferedReader fp) throws IOException

public static String svm_check_parameter(svm_problem prob, svm_parameter param)

public static int svm_check_probability_model(svm_model model)

public static void svm_set_print_string_function(svm_print_interface print_func)