Abstract
In the era of intelligence, we need to carry out continuous autonomous learning and optimization on the data platform, and the first step of continuous autonomous learning is data enhancement. This paper proposes a broad learning method based on cascaded enhancement nodes, which provides a new data enhancement method for continuous autonomous learning on big data platform, and makes it possible for subsequent evolutionary optimization based on learning architecture. Classical broad learning is a typical feedforward neural network, which is not suitable for modeling dynamic time series. In this paper, the feedback structure is introduced into the traditional broad learning system, which makes the enhancement nodes have memory and retain part of the historical information. In the part of feature extraction, phase space reconstruction is used to extract more essential features of the data. At the same time, a weight factor is introduced to assign different weights to each sample according to its contribution to the modeling, eliminate the interference of noise and outliers to the learning process, and improve the robustness of the algorithm. Experimental results show that the proposed algorithm is effective.