Abstract: Optimization-based meta-learning methods for few-shot one-class classification (FS-OCC) aim to fine-tune a meta-trained model to classify the positive and negative samples using only a few ...

Background Annually, 4% of the global population undergoes non-cardiac surgery, with 30% of those patients having at least ...

Offline map generation tailored to GPX routes. Builds compact .mbtiles packages by selecting tiles within a zoom-scaled buffer around your track. Optionally pre-render tiles on a tile server via a ...

Abstract: Multitask learning with a pretext task has excelled in time-series classification task lacking labeled data. The key to multitask learning is to build a pretext task and learn the most ...