Table 1 Comparison of different models.

From: Improving time series forecasting in frequency domain using a multi resolution dual branch mixer with noise insensitive ArcTanLoss

Category

Capturing information at multiple levels

Capturing local information

Capturing global information

Noise robustness

Representative models

RNN-based

 

\(\checkmark\)

  

LSTM22 GRU23

CNN-based

\(\checkmark\)

\(\checkmark\)

  

TimesNet14 MICN24

Transformer-based

  

\(\checkmark\)

 

Autoformer27 PatchTST17

Linear-based

  

\(\checkmark\)

 

DLinear29 TSMixer32

Frequency-based

  

\(\checkmark\)

 

FiLM35 FITS13

Ours

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

FreMixer

Back to article page