Table 1.

Average out-of-sample test errors (rounded average model sparsity in parenthesis) for trac ($a\in \{1,1/2\}$) and sparse log-contrast models, respectively. Each row considers a different base level (OTU, genus, and family). Each number is averaged over ten different training/test splits of the sCD14 data.

Base level	p	trac (a = 1)	trac (a = 1/2)	Sparse log-contrast
OTU	539	6.3e$+$06 (7)	6.7e$+$06 (9)	6.8e$+$06 (8)
Genus	282	6.8e$+$06 (7)	7.1e$+$06 (8)	7.1e$+$06 (9)
Family	112	6.5e$+$06 (4)	6.5e$+$06 (5)	6.6e$+$06 (7)