Treating annotation as a data understanding problem, rather than a labeling workflow challenge, can systematically drive down error rates and reduce the time and cost of producing high-quality data ...

This technical note presents a methodological change to the International Monetary Fund’s Currency Composition of Foreign Exchange Reserves (COFER) dataset. Using a combination of stratified mean ...

The second approach uses generative AI to produce data. Modern generative models are trained on vast amounts of data and can ...

In both cases, it would be better to train the machine learning model with a loss function that ignores the human’s objective and then adjust predictions ex post according to that objective. We ...