12859_2018_2163_MOESM1_ESM.pdf (85.93 kB)

Additional file 1 of Neural networks for link prediction in realistic biomedical graphs: a multi-dimensional evaluation of graph embedding-based approaches

Download (85.93 kB)
journal contribution
posted on 2018-05-21, 05:00 authored by Gamal Crichton, Yufan Guo, Sampo Pyysalo, Anna Korhonen
Results from embedding combination methods explained but not shown in the main manuscript (Hadamard, Weighted-L1 and Weighted-L2). Some analysis on these results. It also contains additional results from the SDNE node embedding creation method explained in the paper. The results in this document also contain the standard deviation from the means reported in the main manuscript. There are also additional Precision at k values for k = 10, 20 and 40. (PDF 74 kb)


Medical Research Council