Table 28 Comparative performance of various models.

From: Air quality prediction based on factor analysis combined with Transformer and CNN-BILSTM-ATTENTION models

Model abbreviation

Factor/principal component

MSE

RMSE

R2

95% Confidence interval

MLFA + Transformer

Factor 1

0.1652

0.4065

0.8231

(− 0.0115, 0.0348)

Factor 2

0.0441

0.2101

0.9597

(0.0003, 0.0464)

PCFA + Transformer

Factor 1

0.6977

0.8353

0.8198

(− 0.0618, 0.0279)

Factor2

0.2579

0.5079

0.9474

(− 0.1256, − 0.0563)

MLFA + N-Beats

Factor 1

0.1912

0.4373

0.7852

(− 0.0746, − 0.0381)

Factor 2

0.1616

0.4021

0.8932

(− 0.0247, 0.0313)

MLFA + Informer

Factor 1

0.2867

0.5354

0.8408

(− 0.1679, − 0.0275)

Factor 2

0.2600

0.5099

0.9433

(0.0806, 0.1149)

Back to article page