Application of unsupervised TSK fuzzy algorithm in large-scale online culture courses

摘要

For the cultural alienation of large-scale online open courses and the matching of key features, a new mapping method is proposed to effectively improve the learning ability of complex nonlinear data algorithms in TSK fuzzy systems. The tendency of cultural alienation is not only reflected in the curriculum knowledge of the text but also in the teaching process. This alienation emerges in the form of cultural obscuration, cultural locking, and cultural decoration. It makes the cultural implication in large-scale online courses gradually reduce, and even goes to the opposite side of culture. Thus, by extracting the convolution layer of the trained VGG19 network model as a feature of a large-scale open curriculum, the proposed features are individually weighted and then connected. The separate features are again convolved and aggregated, and the response function is used to determine connectivity between the open culture courses of the network. Aiming at the problem of high dimensionality of cultural features after single-layer TSK fuzzy feature mapping, the concept of multi-layer progressive fusion is proposed. A fuzzy feature mapping method based on multi-layer progressive fusion is deduced, which effectively solves the data confusion problem caused by excessive feature dimension after mapping. Finally, combined with the unsupervised learning algorithm, the retrieval of large-scale network disordered culture courses is realized. Research shows that the algorithm can effectively identify online culture courses of overlapping scenes without detailed matching process and geometric verification. Compared with the classical fuzzy clustering method, the algorithm has superior and stable performance.