Figure 6 | Scientific Reports

Figure 6

From: Combined unsupervised-supervised machine learning for phenotyping complex diseases with its application to obstructive sleep apnea

Figure 6

Phenotyping and survival prediction for new patients using the trained DPGMM and RSF. Our computational package calculates the cluster assignment probabilities for new patients based on their 43 PSG features (left) by using the trained DPGMM model in this study (Fig. 1b and Table 1). Furthermore, it predicts their comorbidity risks and the corresponding comorbidity-free survival curve (right) by using the trained RSF in this study (Fig. 1c and Fig. 4).

Back to article page