Figure 2 | Scientific Reports

Figure 2

From: Preclinical identification of acute coronary syndrome without high sensitivity troponin assays using machine learning algorithms

Figure 2

Two-dimensional representation of multidimensional study cohort using Gower distance between patient features and UMAP. (A) Dimension reduction by UMAP shows discrete separation into several subclusters. Highlighting patients with MACE (B, red) or ACAO (C, blue) discloses heterogeneous distribution among patient collective and underpins the requirement of supervised machine learning models to identify patients at risk for MACE or ACAO.

Back to article page