Snorkel AI: Putting Data First in ML Development

Re:Snorkel AI: Putting Data First in ML Development by kmax12 on Thursday July 16, 2020 @07:26AM (#23857829)

If anyone is looking for an open source library in this space, I work on one called Compose (https://github.com/FeatureLabs/compose).

With compose, a user defines a labeling function, and then compose scans the historical data looking for training examples to train a machine learning model.

The library has evolved as we apply it to more and more real world use cases, but it is based on approach in this paper from 2016: https://dai.lids.mit.edu/wp-content/uploads/2016/08/07796929....

[ Parent | on HN ]

Re:Snorkel AI: Putting Data First in ML Development by eggie5 on Thursday July 16, 2020 @10:47AM (#23859506)

can you do multi-label w/ Compose? Snorkel only supports single-label.

[ Parent | on HN ]

Re:Snorkel AI: Putting Data First in ML Development by jeff-hernandez on Thursday July 16, 2020 @01:05PM (#23861041)

Yes, you can represent the labeling function as a class and use its methods to represent each label individually.

[ Parent | on HN ]