Predicting Metabolite-Disease Associations Based on Linear Neighborhood Similarity with Improved Bipartite Network Projection Algorithm

Table 1

Parameter analysis for km.

km	10	20	30	40	50
AUCs	0.7598	0.7901	0.8087	0.8370	0.8553

km is changed, and kd is set 10 under 10-fold cross validation.