Skip to content

A recurrent neural network to predict heart failure based on ICD-9 patient data analysis from MIMIC dataset. RETAIN network uses "Reverse Time Attention Mechanism" for predictions.

Notifications You must be signed in to change notification settings

ErniW/RETAIN-heart-failure-prediction-model-RNN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 

Repository files navigation

RETAIN-heart-failure-RNN

A recurrent neural network to predict heart failure based on ICD-9 data analysis. Implementation of RETAIN: An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism

  • It's one of my homework made through Deep Learning for Healthcare course. I've been playing around with code because it introduced a few new things.
  • Its dataset is relatively small for the purpose of the assignment and the training never reached 90% (best result so far is 85%).
  • I can't upload the original notebook due to Coursera Honor Code (Course is a part of MSc in Computer Science). Belive it or not - the code quality there is mediocre and sometimes it's difficult to understand what's going on.
  • There are possible improvements, such as predict stage and probaby more.

About

A recurrent neural network to predict heart failure based on ICD-9 patient data analysis from MIMIC dataset. RETAIN network uses "Reverse Time Attention Mechanism" for predictions.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published