SDMetrics
SDMetrics copied to clipboard
sdv-dev
Metrics to evaluate quality and efficacy of synthetic datasets.
### Environment details * SDMetrics version: 0.21.0 ### Description We have an upcoming fairness metric called [EqualizedOddsImprovment](https://docs.sdv.dev/sdmetrics/metrics/privacy-and-fairness-metrics/equalizedoddsimprovement). This is meant to indicate whether the synthetic data is improving the fairness...
### Environment details * SDMetrics version: 0.21.0 ### Background In certain metrics like [BinaryClassifierPrecisionEfficacy](https://docs.sdv.dev/sdmetrics/metrics/ml-augmentation-metrics/binaryclassifierprecisionefficacy) and [EqualizedOddsImprovement](https://docs.sdv.dev/sdmetrics/metrics/privacy-and-fairness-metrics/equalizedoddsimprovement), the user is generally interested in _augmenting_ the real data with synthetic data. So...
### Problem Description The `DCRBaselineProtection` metric measures the privacy of synthetic data by comparing it against random data. The random data is created by uniformly sampling in the real data's...
### Problem Description While #701 updated the installation instructions needed to use `BNLikelihood` and `BNLogLikelihood` to fix the errors users were encountering using 1.x or 0.14.x versions of pomegranate, it...
### Problem Description Currently, the `DisclosureProtection` metric warns about poor performance when the size of the input data is greater than 50,000 rows. This number was chosen without investigation into...
### Problem Description Right now, the InterRowMSAS metric takes the direct difference between a value in row `n` and row `n+1`. Then, it averages out all these differences. As a...