Overview Frequency Principle/Spectral Bias in Deep Learning

Understanding deep learning is increasingly emergent as it penetrates more and more into industry and science. In recent years, a research line from Fourier analysis sheds light on this magical “black box” by showing a Frequency principle (F-Principle or spectral bias) of the training behavior of deep neural networks (DNNs)—DNNs often fit functions from low to high frequencies during the training. The F-Principle is first demonstrated by one-dimensional (1D) synthetic data followed by the verification in high-dimensional real datasets. A series of works subsequently enhance the validity of the F-Principle. This low-frequency implicit bias reveals the strength of neural networks in learning low-frequency functions as well as its deficiency in learning high-frequency functions. Such understanding inspires the design of DNN-based algorithms in practical problems, explains experimental phenomena emerging in various scenarios, and further advances the study of deep learning from the frequency perspective. Although incomplete, we provide an overview of the F-Principle and propose some open problems for future research.