Skip to content

The repository of using machine learning to analysis xva profile

License

Notifications You must be signed in to change notification settings

cufezhusy/mlXVA

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Introduction

The repository contain the source code for the research project Applying machin learning for troubleshooting credit exposure and XVA profiles conducted in 2018 and 2019.

At this moment, it only contain the Convolutional Neural Network work,

Model training

The model is trained using the GPU platform offered by https://www.floydhub.com/.

Training code job can be found in \jobs

Data used in the training are also uploaded to FloydHub,

One can use the below command to start the model training(it took approx four hours for a standard GPU offered by FloyHub)

floyd run --gpu --env tensorflow-1.11 --data cufezhusy/datasets/gengduoshuju/4:gengduoshuju 'python model_training.py'

Also if you have enough local computational power, can also try to run it locally.

The training results can be found in \train_result

Working paper

Working paper about this project is avaliable in SSRN https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3404863

About

The repository of using machine learning to analysis xva profile

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published