AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |
Back to Blog
Maml meta learning9/28/2023 ![]() Any other differentiable loss functions can be used.Data distribution: q (K samples are drawn from q).→ Find θ that commonly decreases loss of each task Meta-learning is a research field that attempts to equip conventional machine learning architectures with the power to gain meta-knowledge about a range of tasks to solve problems like the one above on a human level of accuracy. On a new task will produce maximally effective behavior. Desired model parameter set is θ such that:Īpplying one (or a small # of) gradient step to θ.Adopted all knowledge-transferable tasks.Task-agnostic (No matter whatever task is).No other assumptions on the form of the model.Classification & regression with differentiable losses, Policy gradient RL.Model-agnostic (No matter whatever model is).Maximizes sensitivity of new task losses to the model parameters. ![]()
0 Comments
Read More
Leave a Reply. |