Support for additional task types and outputs

justin13601 / ACES

ACES: Automatic Cohort Extraction System for Event-Streams

https://eventstreamaces.readthedocs.io/en/latest/

MIT License

21 stars 1 forks source link

Support for additional task types and outputs #53

Open justin13601 opened 3 months ago

justin13601 commented 3 months ago

Right now we basically only support binary classification. This is ok for now, but we could aim to support more complex tasks:

[ ] Ordinal regression
[ ] Multi-class classification
[ ] Multi-label binary classification
[ ] Univariate Regression
[ ] Multivariate Regression
[ ] Sparse Regression
[ ] Sequential decoding prediction tasks
[ ] Time-to-event prediction tasks

Note this overlaps with #72

justin13601 commented 3 months ago

We hope to directly support tasks whose labels cannot be reliably summarized as a single numerical value (ie., a sequential decoding task where we want to predict a sequence of outputs, such as future treatment sequences). For tasks that rely on time-to-event as a paradigm, labels can currently be manually extracted via the window start/end times (ie., subtracting start from end of your window containing your event of interest, and then bucketing appropriately post-hoc).

mmcdermott commented 3 months ago

To-do: merge this and #45