Detailed Experimentation and Comparisons for Continual Learning Methods | HackerNoon
Briefly

The study addresses two significant problems in class-incremental continual learning, emphasizing the need for improved methods that enhance knowledge retention and mitigate forgetting.
In our experiments, we utilized modified experience replay inspired by GDumb, which allows for a balanced representation of classes through effective reservoir sampling.
Read at Hackernoon
[
|
]