Skip to content

interface for transfer learning #11

Description

@lalvim

What about using our own learnt models from problems such as cartpole .... and apply to the initial population for solving another problem?

I think it easy to implement an interface and some heuristic to replicate similar solutions to fill all the pop size.
Sugested Tasks:

  • Design a new method for that

  • Develop an heuristic to do some perturbation on a learnt model (e.g. cartpole and others) in order to fill the population of hundreds of initial solutions

  • Create an experiment on tinymlzoo

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Type

    No type

    Fields

    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions