Skip to content
/ MRNN Public
forked from jsyoon0823/MRNN

Multi-directional Recurrent Neural Networks

Notifications You must be signed in to change notification settings

nbharaths/MRNN

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

23 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

MRNN

Multi-directional Recurrent Neural Networks

  1. Datasets (GOOGLE.csv, GOOGLE_Missing.csv)
  • These datasets are the example time-series datasets that can be used for testing M-RNN Architecture with Data_Loader.py
  1. Data_Loader.py
  • Using the Raw datasets, it extracts the features and time information.

  • It also divides training and testing sets for further experiments

  • It has two input parameters (1). train_rate: training / testing set ratio (2). missing_rate: the amount of introducing missingness

  • It has 4 outputs for each training and testing set (1). X: Original Feature (2). Z: Feature with Missing (3). M: Missing Matrix (4). T: Time Gap

  1. M_RNN.py
  • Using the outputs of the Data_Loader.py, it imputes the missing features using M-RNN architecture
  • It consists of Bi-directional GRU and MLP.
  • The details of the M-RNN architecture can be found in the following link.
  • https://arxiv.org/pdf/1711.08742.pdf
  1. M_RNN_Main.py
  • Combine the above three components with MSE performance metric.

About

Multi-directional Recurrent Neural Networks

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%