jmlr.org/papers/volume23/21-0308/21-0308.pdf

• Learning optimisers for different problems. trying to improve over the hard engineered optimisation algos like sgd , adam,etc.

model based methods