Iterative Teaching by Label Synthesis
arXiv:2110.14432
Abstract
In this paper, we consider the problem of iterative machine teaching, where a teacher provides examples sequentially based on the current iterative learner. In contrast to previous methods that have to scan over the entire pool and select teaching examples from it in each iteration, we propose a label synthesis teaching framework where the teacher randomly selects input teaching examples (e.g., images) and then synthesizes suitable outputs (e.g., labels) for them. We show that this framework can avoid costly example selection while still provably achieving exponential teachability. We propose multiple novel teaching algorithms in this framework. Finally, we empirically demonstrate the value of our framework.
NeurIPS 2021 Spotlight (v5: 28 pages, 20 figures, fixed typos in v4)
References in corpus (8)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Distilling the Knowledge in a Neural Network
- An Overview of Machine Teaching
- Near-Optimally Teaching the Crowd to Classify
- Policy Poisoning in Batch Reinforcement Learning and Control
- Training Set Debugging Using Trusted Items
- Preference-Based Batch and Sequential Teaching: Towards a Unified View of Models
- Neural Similarity Learning