WUR-AI / WUR-AI/diffWOFOST

[Backlog Item]: Build hybrid modeling framework

Open
#11 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Backlog item Epic 1
Dominant language
Python
Stars
25
Forks
6
Avg merge
2d 23h
Merged PRs (30d)
1

Description

Based on the prototype in #3, design and implement the basic functionalities of the hybrid modeling framework.

Explanation:
The current wofost72_pp consists of several crop modules. Each crop module is a physics based simulation. In hybrid modeling, each module can be in principle replaced with a differentiable mathematical simulation where the parameters can be learned in training, see explanations in #3 and this comment. Note that in every tests module where we check the differentiability, the physics based simulation is wrapped with torch.nn.module.

Requirements:

  • uses torch.nn.Module
  • support mix and match of ML methods
  • supports training and optimization
  • support different loss functions

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the prototype in issue #3, the explanation linked from issue #82, and the existing differentiability tests that wrap physics-based simulations with torch.nn.Module. Define the framework boundaries before implementing the requirements for interchangeable ML methods, training and optimization, and multiple loss functions. Done means these capabilities are demonstrated by tests or examples.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.