Skip to Main Content
Table 9 
Analysis of refinement methods applied to the same initial mappings of our adversarial autoencoder on the Dinu-Artetxe data set.
 En-ItEn-EsEn-DeEn-Fi
Artetxe Refinement 
Robust self-learning 44.5 40.5 36.5 30.6 46.8 42.9 31.5 30.4 
Robust self-learning + Symmetric re-weighting 47.9 42.6 37.5 32.1 47.9 44.1 32.9 33.0 
  
Our Refinement 
Procrustes solution 45.3 39.4 35.2 29.9 46.8 42.6 30.4 31.9 
Symmetric re-weighting 46.5 42.4 37.5 31.9 48.3 44.1 32.4 32.7 
Procrustes Solution + Symmetric re-weighting 47.7 42.3 38.1 32.3 48.7 44.1 32.6 33.2 
 En-ItEn-EsEn-DeEn-Fi
Artetxe Refinement 
Robust self-learning 44.5 40.5 36.5 30.6 46.8 42.9 31.5 30.4 
Robust self-learning + Symmetric re-weighting 47.9 42.6 37.5 32.1 47.9 44.1 32.9 33.0 
  
Our Refinement 
Procrustes solution 45.3 39.4 35.2 29.9 46.8 42.6 30.4 31.9 
Symmetric re-weighting 46.5 42.4 37.5 31.9 48.3 44.1 32.4 32.7 
Procrustes Solution + Symmetric re-weighting 47.7 42.3 38.1 32.3 48.7 44.1 32.6 33.2 
Close Modal

or Create an Account

Close Modal
Close Modal