Skip to main content

Table 4 Prediction accuracies of LDA, modified KNN and LGBM classification algorithms at different time segments for the best combination of HRV parameters

From: Prediction and detection of virtual reality induced cybersickness: a spiking neural network approach using spatiotemporal EEG brain data and heart rate variability

ECG prediction

Time segment

ML algorithm

LDA

KNN

LGBM

2 Min baseline

56.5% SI + SDNN

74.2% PNS + SNS

69.4% SNS

15–45 s

61.3% PNS + SNS + SDNN + RMSSD

67.7% SI

67.7% SNS + SI + RMSSD

75–105 s

62.9% PNS + SNS

74.2% SNS

71.0% SNS + SDNN

25–35 s

6.5% RMSSD

67.7% RMSSD

61.3% RMSSD

85–95 s

16.1% RMSSD

59.7% RMSSD

51.6% RMSSD

  1. Top accuracies are highlighted in bold