Skip to content

Latest commit

 

History

History
30 lines (18 loc) · 1.01 KB

README.md

File metadata and controls

30 lines (18 loc) · 1.01 KB

dreem

Challenge Dreem

/!\ Please make sure the folder data actually contains the following files: /!
- quality_dataset.h5 - record1.h5 - record2.h5 - record1_bad_quality.csv - record1_good_quality.csv

/!\ Code is to be run with src folder as the working directory. /!\

Files descriptions :

  • easy_import.py: contains extract_signals function to load record1.h5 and record2.h5.

  • exploratory.py: exploratory analysis of quality_dataset.h5.

  • functions.py: a bunch of useful functions.

  • quality_detector.py: implementation of the final model to predict signal quality on record1.h5 and record2.h5.

  • randomForest.py: file to build a Random Forest classifier unsing quality_dataset.h5.

  • incrementsVarRF.py: file to build a Random Forest classifier unsing quality_dataset.h5 after extraction piece-wise variances as features.

  • RFacf.py: file to build a Random Forest classifier unsing quality_dataset.h5 after extraction piece-wise variances and ACF (Auto Correlation Functions) as features.