Chapter 17 – Reinforcement Learning Frontiers
- Meta learning produces a versatile AI model that can learn to perform various tasks without having to train them from scratch. We train our meta-learning model on various related tasks with a few data points, so for a new but related task, it can make use of the learning obtained from the previous tasks and we don't have to train it from scratch.
- Model-Agnostic Meta Learning (MAML) is one of the most popularly used meta-learning algorithms and it has created a major breakthrough in meta-learning research. The basic idea of MAML is to find a better initial model parameter so that with good initial parameters, the model can learn quickly on new tasks with fewer gradient steps.
- In the outer loop of MAML, we update the model parameter as and it is known as a meta objective.
- The meta training set basically acts as a training set in the outer loop and is used to update the model parameter in the outer...