Table 2.

Comparison of the results ${\mathsf{p}}_{\text{tensor}}$ of Algorithm 1 based on low-rank tensors and ${\mathsf{p}}_{\text{vector}}$ of (13) based on matrix operations for different Mutual Hazard Networks from Schill et al. (2019)

Disease	d	Relative residual ${\mathsf{p}}_{\text{vector}}$	Relative residual ${\mathsf{p}}_{\text{tensor}}$	Relative Euclidean distance	Absolute KL divergence	Storage cost ${\mathsf{p}}_{\text{vector}}$	Storage cost ${\mathsf{p}}_{\text{tensor}}$
Breast cancer	10	1.7e$-$17	8.8e$-$05	3.3e$-$04	3.3e$-$07	1024	3052
Colorectal cancer	11	2.9e$-$17	9.8e$-$05	3.5e$-$04	3.1e$-$07	2048	4396
Renal cell carcinoma	12	1.1e$-$17	9.8e$-$05	1.2e$-$04	1.4e$-$06	4096	4553
Glioblastoma	20	1.6e$-$17	9.7e$-$05	9.7e$-$05	4.4e$-$06	1,048,576	44,065