Table 1 The weighted KM-SVM (or SVM) with the boosting algorithm

1. Initialize the weight \(c_n\) with \(\frac{1}{N}\).

2. For \(m=1\) to M:

  (1) Fit a KM-SVM (or SVM) \(G_m(x)\) with weights \(c_n\) to clustering centers of train data.

  (2) Compute

    \(err_m = \frac{\sum _{n=1}^N c_n I \big (y_n \ne G_m(x_n) \big ) }{\sum _{n=1}^N w_n }\)

  (3) Compute \(\alpha _m = \text {log} \big ( \frac{1-err_m}{ err_m } \big )\)

  (4) Set \(c_n \leftarrow c_n \cdot \text {exp} \big [ \alpha _m \cdot I (y_n \ne G_m(x_n) ) \big ]\)

3. Output \(G(x) = \text {Sign}\Big [ \sum _{m=1}^M \alpha _m G_m(x) \Big ]\).