techniques for adversarial tasks

[tliu68]

Is your feature request related to a problem? Please describe.
Those PRs/issues (#79, #3, #389, #394) are trying to improve the performance of progressive learning in adversarial settings for different kinds of data or ??.

Describe the solution you'd like
We (me and @williamrxu @ykobaya9 @latasianguy) would like to document techniques for overcoming difficulties in a progressive learner’s performance with different types of adversarial tasks.