Description

DSpace is open source software for deep learning programming

Demonstration video

Primary features

• Open source, in memory scalable machine learning

• Implemented on H2O's distributed Map/Reduce framework

• Data parser that guesses the schema of incoming datasets

• Supports data ingestion from multiple sources in various formats

Example use cases

• Chicago Crime Prediction

• Airlines delays prediction

• Ham or Spam

More examples at http://docs.h2o.ai/h2o/latest-stable/index.html

Community

• Blog https://blog.h2o.ai/

See also

• https://www.h2o.ai/

DSpace in Labs Workbench

Basic instructions for using H2O in Labs Workbench

• Add the H2O AI application

  • Select Launch

  • On the Labs Workbench Applications view, expand the "H2O" application and click "View Config" beside its "H2O" service

  • You should see a generated value for ADMIN_PASSWD

  • Once the application has started, expand the "DSpace" application and click the DSpace endpoint link

  • You should then be able to log into H2O using your Labs Workbench account's e-mail and the generated password