Skip to content

CoDaS-Lab/self-teaching

Repository files navigation

Self-teaching

Code associated with:

Yang, S. C-H., Vong, W.K., Yu, Y., & Shafto, P. A unifying computational framework for teaching and active learning (submitted)

Directory structure

  • models: Contains the different models (active learning, teaching and self-teaching) used in the simulations
  • notebooks: Contains various jupyter notebooks with worked examples
  • old_models: Original matlab code of Pat's original pedagogical sampling model for the causal graph task and unused models
  • simulations: Code to run simulations (currently only self-teaching for the concept learning task)
  • tests: Directory for test code
  • run_simulations.py: Main file to run the simulations and generate the figures from the paper

Running the code

To run the simulations and produce the figures in the paper:

python run_simulations.py

To run the tests:

pytest

The tests require pytest on your machine, which can be installed with the following:

pip install -U pytest

About

Formulating active learning as self-teaching

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published