Skip to content

Latest commit

 

History

History
executable file
·
38 lines (27 loc) · 1.04 KB

README.md

File metadata and controls

executable file
·
38 lines (27 loc) · 1.04 KB

ANYTIME ACTIVE LEARNING

This repository contains the code used for experiments in Anytime active learning research published on the following paper:

Ramirez-Loaiza, Maria E., Aron Culotta, and Mustafa Bilgic. 2014. “Anytime Active Learning.” In AAAI Conference on Artificial Intelligence.

Content

In this file:

  • Repository
  • Configuration
  • How to run

Repository

Structure of the repository

  • datautil -> various functions to load the data
  • experiment
    • |_ anytime.py --> active learning loop for anytime AL
    • |_ experiment_utils.py --> printing resutls
  • strategy
    • |_ baselearner.py --> basic student class
    • |_ randomsampling.py --> AnytimeLearner
    • |_ structured.py --> AnytimeLearner with structured reading classes
  • expert
    • |_ baseexpert.py --> you need the class NeutralityExpert
  • sentence
    • |_ baseexpert.py --> you need the class NeutralityExpert
  • scripts --> contains all scripts used to run experiments for this research

Configuration

Coming soon.

How to run an experiment

Coming soon.