Looking at the “Cumulative Gain Chart”, in the first 0.5% of the population, the predictive model is 6 times more efficient compared to segmentation (10% of respondents selected by predictive ...

For that second chunk of data, we can apply AI models to further segment the 100-terabyte chunk into buckets based on the expected probability of a file having PII.

While deep learning-based segmentation methods have demonstrated state-of-the-art performance, they often rely on vast amounts of labeled data, which is expensive and time-consuming to obtain.

More information: Qingyuan He et al, Exploring Unlabeled Data in Multiple Aspects for Semi-Supervised MRI Segmentation, Health Data Science (2024). DOI: 10.34133/hds.0166 ...