Abstract

Analyzing the differences in the development of advertising education between Chinese and Thai universities can bring insights into the development direction of advertising education in China. In this article, we examine the theoretical foundations of the single hidden layer feedback neural network-based extreme learning machine algorithm. To address the challenge of SLFNs' lengthy training times, iteration is used to continuously adjust the input weights and bias vectors of the hidden units. Meanwhile, due to the poor generalization ability of the ELM algorithm, the Mercer condition is used to construct the kernel matrix instead of the output matrix to establish the ELM model with kernel function, and the difference in the development of college advertising education between China and Thailand is analyzed based on the KELM algorithm.From 2011 to 2018, the percentage of advertising practitioners with education at bachelor's degree and above in China has increased by 14.5%, and that in Thailand is 17.6%. The teacher-student ratio of advertising education in Chinese universities has improved by nearly 700% from 2001 to 2020, while Thailand has improved by nearly 3000%. Meanwhile, the overall evaluation index of advertising teaching in Thai universities has improved by 22.93% since entering the new century, and China has also improved by 19.84%. Based on ELM's analysis of the real situation and developmental differences of advertising education in each country in the new era, it can provide a perspective for observation and consideration of college advertising education, which is expected to be reformed.