• Complete workover.
• New functions such as parallel.
• Namespace %SYS.

Initial Release

• Introducing query for random sample
• Models performance improvement, by random sampling train and test datasets (thanks @Raj.Singh5479 for this tip)
  • No-show prediction: from 81% to 95; with great improvement in recall rate (from 11% to 83%)
  • Heart failure death risk: from 89% to 98%
• Documentation update

• increased performance of charts with big data sets

• fixed issue with MultiPolygon in GeoJSON support (polygons with holes)

Updated escape sequence color scheme

ProcessOneAxisCell: populate axis cell info after determining cell visibility

• added "View As" options: "Stacked bar chart", "Stacked column chart" (#179)
• fixed formatting for charts (#185)
• added percentage display for pie chart (#187)
• added annotations display for pie chart (#186)

• New model for Heart Failure prediction
• Swagger API documentation
• Documentation update

Added new metric: BarcodeSimilarity

New logo.
Added logistic regression process to compare with linear regression's getting coefficients model.

• Add charts with prediction
• Move all config commands from jupyter notebook to dockerfile

Initial Release

Index rebuilding added in module deployment

• Fix no-show model training method when there is no model previouslly trained
• Error handling in no-show detection service
• Updates on postman collection
• Code clean up
• Doc updates

Added support zpm+hspm version

Added calculation of prediction accuracy

Don't ask for parameters if not in terminal mode

• Model example for no-show detection on appointments, trained on FHIR resources
• DTL transformation for using FHIR resources on IntegratedML
• API for predicting no-show in FHIR resources

Added index on code2. Fixed some data issues

ssl-client dependency fix - now works with the latest ssl-client

Initial Release

Initial Release

Initial Release

Right description.

improvements and bug fixes

Added description to module.xml

Added Docker support

Adjustments for zpm explorer web app

Improved IRIS BI export - ##class(dev.code).export("*.DFI")