Frequency-domain MLPs are More Effective Learners in Time Series Forecasting.
Kun YiQi ZhangWei FanShoujin WangPengyang WangHui HeDefu LianNing AnLongbing CaoZhendong NiuPublished in: CoRR (2023)
Keyphrases
- frequency domain
- spatial domain
- fourier transform
- power spectrum
- cross correlation
- denoising
- feature extraction
- fourier analysis
- facial asymmetry
- frequency domain analysis
- frequency analysis
- low pass
- artificial neural networks
- phase correlation
- power spectra
- neural network
- discrete fourier transform
- fast fourier transform
- high quality
- spectrum analysis
- affine invariance
- image segmentation