is it possible to do continual training with an auto_encoder?

[andrewm4894]

lets say i want to continually train my detector on the recent week of data.

is there a way in the pyod api to actually have a auto_encoder detector ans just train it continually on each new batch of data each week?

i dont want to have to make a big training job where we start from scratch each week and train on say last two months of data.

instead i'd like to just continually train on the recent week of data and update my model that way.

i know this might not be most stable but just wondering if this is possible with the current pyod api?