[NEW] Khan et al. "Characterization of Term and Preterm Deliveries using Electrohysterograms Signatures"

[GillesVandewiele]

Reported result

The system achieves 95.5% accuracy on publicly available Term-Preterm EHG Database.

Features

In this research four type of features are extracted from the EHG signatures such as; Median frequency [33], Shannon energy [34], Log energy [35], Lyapunov exponent [36] for the categorization of the EHG waveforms.

Model

This research uses support vector machine (SVM)

Oversampling

In this research, adaptive synthetic sampling approach (ADASYN) [31, 37, 38]is used

Comments

[GillesVandewiele]

It seems that the Lyapunov exponents are missing from the raw_features.csv file. For this, the slow parameter of FeaturesJager has to be set to true. It can take over 24 hours to extract the features for all 300 signals then though...

[gykovacs]

I think we can replace Lyapunov with Higushi, it should change much.

[GillesVandewiele]

I still have the old output which contains the lyapunov exponents... Perhaps we could merge that into our raw_features file?

EDIT: It also contains the Correlation dimension which is used in #29