Fig. 3: Outlier detection effect of ODBAE is generally better than other schemes. | Communications Biology

Fig. 3: Outlier detection effect of ODBAE is generally better than other schemes.

From: ODBAE: a high-performance model identifying complex phenotypes in high-dimensional biological datasets

Fig. 3: Outlier detection effect of ODBAE is generally better than other schemes.The alternative text for this image may have been generated using AI.

a–c Data point distribution plot for 3 synthetic dataset, including 2-dimensional Gaussian distribution dataset with diagonal covariance matrix (a), 2-dimensional Gaussian distribution dataset with non-diagonal covariance matrix (b) and 2-dimensional Gaussian distribution dataset with diagonal covariance matrix and non-Gaussian distributed noise (c). d–f AUC scores of comparison results of dataset in this figure (a, d), (b, e), and (c, f). g–i AP scores of comparison results of dataset in this figure (a, g), (b, h), and (c, i). j, k Comparison results for Dry Bean Dataset. l, m Comparison results for Breast Cancer Dataset. In (d–m), all methods were run 10 times under different outlier ratios, and error bars indicate mean ± standard error.

Back to article page