Skip to content

300 features selection, CatBoost, NannyML to monitor concept drift, auto-retrain, daily load h5 trick

Notifications You must be signed in to change notification settings

GrigoriiTarasov/Optiver-Trading-at-the-Close-retrain-monitoring

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

28 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Optiver-Trading-at-the-Close-retrain-monitoring

Implementation of top solutions of Optiver - Trading at the Close 2023 with
ML engineering best practices extension

Stage Status
1 EDA w/ target as stock wap regression restore
2 Feature generation, 370+ total
3 Batch learning on daily h5 data
4 Feature selection with Catboost
5 Params optimization with Optuna and MlFlow
on selected features
6 NannyMl to check perfomance decay for ts splits
7 Retrain options and DVC each fited porion

About

300 features selection, CatBoost, NannyML to monitor concept drift, auto-retrain, daily load h5 trick

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published