Risk Prediction of Diabetic Nephropathy via Interpretable Feature Extraction from EHR Using Convolutional Autoencoder.
Takayuki KatsukiMasaki OnoAkira KosekiMichiharu KudoKyoichi HaidaJun KurodaMasaki MakinoRyosuke YanagiyaAtsushi SuzukiPublished in: MIE (2018)
Keyphrases
- feature extraction
- diabetes mellitus
- risk factors
- prediction accuracy
- early warning
- diabetic patients
- risk management
- restricted boltzmann machine
- risk assessment
- feature selection
- prediction model
- prediction error
- discriminant analysis
- high risk
- face recognition
- preprocessing
- feature vectors
- wavelet transform
- medical data
- neural network
- image processing
- prediction algorithm
- artificial neural networks
- multiresolution
- pattern classification
- texture analysis
- frequency domain