ML on Rails is designed to simplify the process of complex machine learning model creation, training, and validation, and analysis, allowing uses the opportunity to quickly explore and analyze data. An online wizard quickly assists the user in creating a ML model, then continues model development by stepping users through training and validation ensuring models achieve optimal accuracy. With a direct connection to our enterprise Hyperion Data Platform providing data interoperability and management, it is easy to select and build unique healthcare datasets or upload your own data all within an analytical and business intelligence environment with a familiar graphical user interface.
Features:
- Built upon our custom enterprise Hyperion Data Platform providing data interoperability and management, with data harmonization through our Hyperion Common Data Model.
- Data-viewer to explore Hyperion data and build datasets:
- Auto detection of data type
- Hints for machine learning model development
- On-premise secure environment maintaining data security and privacy.
- Wizard to quickly create models:
- Multiple variations of models are supported (LSTM, RNN, CNN, etc.)
- Features familiar drag-and-drop interface
- Explanations of common ML terms & parameters
- Suggestions for optimal hyperparameters
- Automated training and evaluation
- Runs on Hyperion’s state of the art computing infrastructure
- Export options to download and share models with others
- Advanced configuration options for experienced ML developers