Skip to main content

Table 3 Configuration of the forecasting models for wellness conditions

From: Forecasting one-day-forward wellness conditions for community-dwelling elderly with single lead short electrocardiogram signals

Models

Optimized superior parameters

Deep learning-based

LSTM

h: 256, i: 100, η0: 0.6, f: 0.1. Optimized scheme of η: η·f when ε not improving every 10 epochs.

 

BiLSTM

h: 256, i: 100, η0: 0.6, f: 0.1. Optimized scheme of η: η·f when ε not improving every 10 epochs.

Machine learning-based

ANN

h: 100, η0: 0.01, f: 0.01. Optimized scheme of η: η·f per 100 epochs

 

SVM

C: 10, Γ: 0.01, Kernel: RBF Tolerance for stopping criterion: 10−3