Method SVM_predict_probability

svm.go:2356–2393

(model *SVM_Model, x []SVM_Node, prob_estimates []float64)

}

func (this *SVM) SVM_predict_probability(model *SVM_Model, x []SVM_Node, prob_estimates []float64) float64 {
	var rst float64
	if (model.param.svm_type == C_SVC ||
		model.param.svm_type == NU_SVC) &&
		model.probA != nil && model.probB != nil {
		var i int
		nr_class := model.nr_class
		dec_values := make([]float64, nr_class*(nr_class-1)/2)
		this.SVM_predict_values(model, x, dec_values)

		min_prob := float64(1e-7)
		pairwise_prob := make([][]float64, nr_class)
		for i = range pairwise_prob {
			pairwise_prob[i] = make([]float64, nr_class)
		}

		k := 0
		for i = 0; i < nr_class; i++ {
			for j := i + 1; j < nr_class; j++ {
				pairwise_prob[i][j] = math.Min(math.Max(this.sigmoid_predict(dec_values[k], model.probA[k], model.probB[k]), min_prob), 1-min_prob)
				pairwise_prob[j][i] = 1 - pairwise_prob[i][j]
				k++
			}
		}
		this.multiclass_probability(nr_class, pairwise_prob, prob_estimates)

		prob_max_idx := 0
		for i = 1; i < nr_class; i++ {
			if prob_estimates[i] > prob_estimates[prob_max_idx] {
				prob_max_idx = i
			}
		}
		rst = float64(model.label[prob_max_idx])
	} else {
		rst = this.SVM_predict(model, x)
	}
	return rst
}

func (this *SVM) SVM_save_model(model_file_name string, model *SVM_Model) {