What is the advantage of adversarial Loss? about gain HOT 5 CLOSED

Adversarial loss effects less in missing completely at random.
When the missingness comes from missing at random or missing not at random settings, the effects of adversarial loss increase.
You can try different datasets with different missingness settings.

from gain.

Well noticed. Thank you very much for your reply!

from gain.

Dear all, please bear me for adding question after this issue was closed, but I think my question is relevant so I post it here.
I was wondering that Proposition 2 requires that M and X are independent? How could the theoretical analysis be adapted to the missing not at random mechanism (MNAR)?
In table 3 of supplementary document, indeed we can see that GAIN is much better than auto-encoder under MNAR. Does the implementation of auto-encoder use M as additional input? Or it is a simple implementation with only X as input?

from gain.

1. We only prove the theoretical works in missing completely at random setting.

• Thus, it is not directly adapted to MNAR and MAR settings.

2. We use M as the additional inputs for MNAR and MAR settings that we would like to capture the information in the mask vector.

from gain.

Thanks for your prompt reply. From the results, it is a remarkable feature of GAIN for handling MNAR or MAR!

from gain.