why use multiplication in TransE model #12
Description
In transE model, the loss of a triple is usually defined as (h + r - t), but in the code, it uses multiplication. May I know why?
{{ message }}
In transE model, the loss of a triple is usually defined as (h + r - t), but in the code, it uses multiplication. May I know why?