arXiv (Cornell University)· 2015· Preprint
Distilling the Knowledge in a Neural Network
- 13,986citations
- 2015year
Short summary
A novel knowledge distillation technique compresses an ensemble of neural networks into a single, deployable model, achieving surprising results on MNIST and improving a commercial acoustic system.
AI-generated from the title and abstract; the full text is not read.
Key points
- Knowledge distillation can compress large neural network ensembles into a single, deployable model.
- The technique achieved surprising results on the MNIST dataset.
- A heavily used commercial acoustic system was significantly improved by distilling ensemble knowledge.
- A new ensemble type uses specialist models to distinguish fine-grained classes confused by full models.
AI-generated from the title and abstract; the full text is not read.
Abstract
A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions. Unfortunately, making predictions using a whole ensemble of models is cumbersome and may be too computationally expensive to allow deployment to a large number of users, especially if the individual models are large neural nets. Caruana and his collaborators have shown that it is possible to compress the knowledge in an ensemble into a single model which is much easier to deploy and we develop this approach further using a different compression technique. We achieve some surprising results on MNIST and we show that we can significantly improve the acoustic model of a heavily used commercial system by distilling the knowledge in an ensemble of models into a single model. We also introduce a new type of ensemble composed of one or more full models and many specialist models which learn to distinguish fine-grained classes that the full models confuse. Unlike a mixture of experts, these specialist models can be trained rapidly and in parallel.
The authors' abstract, as published at the source. arXiv (Cornell University), 2015 · DOI ↗
Continue with a free account
Ask the paper: 3 free questions a day about this paper; save it, get its citation, new summaries every day for your field. Takeaways are Premium.
Continue free on the webSign in with Google or Apple; no card needed. You come back to this paper.
On your phone:
Field: Artificial Intelligence
Artificial IntelligenceComputer Science