Cloud Data System Design for Mental Crisis State Recognition of College Students Based on Machine Learning

Table 4

Performance comparison between the research work in this article and existing related emotion recognition research.

Method	Cross individual verification	Result
QTFD and SVM	NO	85.9% (valences, 2 classes)
Statistical, power spectral and HOC features, and SVM	NO	58.9% (valences, 3 classes)
Entropy, PS features, LSP, KNN, SVM, and MLP	NO	63.48% (valences, 3 classes)
DE, DASM, RASM, and SVM	NO	83.98% (3 classes)
DE and SVM	Yes	60.94% (3 classes)
ApEn, PerEn, ShEn, and SVM	Yes	83.34% (2 classes)
DySampEns and SVM	Yes	85.12% (2 classes)
		64.16% (3 classes)