Ridge Regression Based on Glowworm Swarm Optimization Algorithm with t-Distribution Parameters

With the continuous development of machine learning and artificial intelligence (AI) related algorithms and methodologies, ridge regression, which is an important representative regression method in machine learning, can effectively overcome the problem of solutions with some items irreversible in multiple linear regression by adding L2 regularization term. Moreover, multiple linear regression have its own characteristics or constraints like cumbersome calculations, sparse, unstructured data, etc. In this paper, an ridge regression based on glowworm swarm optimization algorithm with t-distribution parameters has been proposed, which has a higher calculation efficiency in obtaining the optimal solution, and can effectively help the solution to avoid from falling into the local minimum trap.