Investigation into art education course suggestion algorithms using a lightweight gradient boosting machine and deep belief network

The rapid development of artificial intelligence has not only profoundly changed all walks of life, but also had a profound impact on the field of education. Schools across the country are also actively organizing and carrying out online teaching and research activities through various online teaching platforms. However, the current course recommendation method has a certain knowledge lag and the computational efficiency needs to be improved. In response to the above problems, this paper designs a feature extraction method based on a deep belief network, and designs an intelligent aesthetic education course recommendation algorithm in combination with an improved lightweight gradient boosting machine. The research results show that the research method only needs 20 iterations, and the corresponding loss values of the training set and the validation set converge to 0.23 and 0.08 respectively. Compared with the current latest algorithm, the accuracy of the research method is 98.6% when it is stable, and the calculation time is at least 6.799s/epoch, and the F1 value is as high as 86.62%. The above research results show that the research method can not only improve the efficiency of acquiring aesthetic education knowledge content, but also promote the diversified channels for receiving knowledge and flexible learning methods.