Locally Linear Embedding Based on Seagull Optimization Algorithm with T-distribution Parameters

With the development of machine learning and artificial intelligence, manifold structure is widely studied because it can extract relatively simple structure from high dimensional data. Locally linear embedding (LLE) has attracted more and more attention with the idea of algebraic structure. However, due to the constraints of locally linear embedding, there are computational problems such as tedious, sparse and unstructured. In this paper, we propose a seagull optimization algorithm with t-distribution parameters for global optimizer, which avoids the trap of falling into local minima and can simply and accurately find the optimal parameters.