Skip to content

A platform-centric approach to developing analytical tools and solutions aiming to maximize the advantages of combining multiple analytic toolsets (AI/ML, BI, SQL, Quantum, IoT, MapReduce, NLP, DL/CV, etc.) in a single analytic process or a system of agent processes.

License

Notifications You must be signed in to change notification settings

lukyanchikov/Convergent-Analytics

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

This repo is an archive. Keep following us here.

Featured Materials

Breaking News

OpenEMR (the open-source Electronic Medical Record system) - now extended with Stroke Prediction solution interface whose logic is deployed in jBPM (the open-source orchestration engine). Production-level Health AI is here and is for anyone. Follow the related post series here.

Stroke Prediction user interface in OpenEMR

jBPM orchestrating Python. Open source from A to Z, zero customization required to begin. Follow the related post series here.

jBPM orchestrating Python

Advanced Solution Features Developed Using Convergent Analytics ML Toolkit Framework:

Model Training Navigator (MTN):

  • Learning from history of the accuracy metrics resulting from model testing - and suggesting (using GAN, generative adversarial networks) hyperparameters for model training (at a given level of minimum expected accuracy):

Generated Hyperparameter Substitution

  • Achieving the expected accuracy using MTN in No-Show Prevention solution:

Expected Accuracy Achieved in No-Shows

  • Achieving the expected accuracy using MTN in Readmission Prevention solution:

Expected Accuracy Achieved in Readmission

Solutions Developed Using Convergent Analytics ML Toolkit Framework:

Healthcare AI:

  • No-Show Prevention

Automates no-show probability estimation on a real-time flow of appointments.

No-Show Prevention

  • Readmission Prediction

Automates readmission probability estimation on a real-time flow of clinical episodes.

Readmission Prediction

Webinars

Articles

Collaboration Experiences

Tutorials

Demos

Links to Required Downloads

  • InterSystems IRIS Community Edition: download a community edition of InterSystems IRIS data platform here - create an account to start the download
  • ML Toolkit: Python Gateway: download an open-source set of extensions to InterSystems IRIS for in-platform orchestration of Python here
  • ML Toolkit: R Gateway: download an open-source set of extensions to InterSystems IRIS for in-platform orchestration of R here
  • ML Toolkit: Julia Gateway: download an open-source set of extensions to InterSystems IRIS for in-platform orchestration of Julia here

Repository Resources Legend

  • Programs: program classes for pre-configured showcases
  • Data: source data for showcases
  • Tables: table classes for showcases

Root Resources

Showcases (Programs/Data/Tables):

  • 000 Robotized ML - FACTORY.xml/FACTORY.zip/FACTORY.txt
  • 001 Sentiment Analysis - SENTIMENT.cls/SENTIMENT_NEG_UTF8.txt, SENTIMENT_POS_UTF8.txt, SENTIMENT_UNK_UTF8.txt/SENTIMENTNEG.cls, SENTIMENTPOS.cls, SENTIMENTUNK.cls
  • 002 Engine Condition Classification - ENGINES.cls/ENGINES_EVEN.txt, ENGINES_ODD.txt/ENGINESEVEN.cls, ENGINESODD.cls
  • 003 Reimbursement Request Check - RECORD.cls/RECORD_TEST.txt, RECORD_TRAIN.txt/RECORDTEST.cls, RECORDTRAIN.cls
  • 004 Retail Cannibalization Analysis - CANNIBALIZATION.cls/CANNIBALIZATION.txt/CANNIBALIZATIONVOLUMEYEARWEEKCATEGORYDESC.cls
  • 005 Marketing Campaign Optimization - CAMPAIGN.cls/CAMPAIGN_APPLY.txt, CAMPAIGN_TRAIN.txt/CAMPAIGNAPPLY.cls, CAMPAIGNTRAIN.cls
  • 006 Rail Time Series Discovery - INTRA.cls/INTRA.txt
  • 007 Housing Debts Prediction - HOUSING.cls/HOUSING_TEST.txt, HOUSING_TRAIN.txt/HOUSINGTEST.cls, HOUSINGTRAIN.cls
  • 008 Diseases Network Analysis - DISEASES.cls/DISEASES_MATRIX.txt, DISEASES_NODES.txt/DISEASESMATRIX.cls, DISEASESNODES.cls

About

A platform-centric approach to developing analytical tools and solutions aiming to maximize the advantages of combining multiple analytic toolsets (AI/ML, BI, SQL, Quantum, IoT, MapReduce, NLP, DL/CV, etc.) in a single analytic process or a system of agent processes.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published