 |
ML Newsnew | past | best | rss | submit | about |
| Towards Transparent Machine Learning (arxiv.org) | |
| 2 points by nwah1 on Nov 18, 2019 | hide | past | pdf | discuss on HN | |
|
In plain words: Instead of a black box, this approach writes the trained model and the training process as source code, so people can read, check, and edit them like any program. The paper argues this would make AI systems safer and more secure than opaque models today. Abstract · Position Paper: Towards Transparent Machine LearningTransparent machine learning is introduced as an alternative form of machine learning, where both the model and the learning system are represented in source code form. The goal of this project is to enable direct human understanding of machine learning models, giving us the ability to learn, verify, and refine them as programs. If solved, this technology could represent a best-case scenario for the safety and security of AI systems going forward. | |
| add comment on HN |