Hybrid GA-SQP neural network frame for solving english language learning

用于解决英语语言学习问题的混合 GA-SQP 神经网络框架

阅读:1

Abstract

A new way to compute the English Language Mathematical Model (ELMM) is described in this paper using a neural network hybridized with a genetic algorithm-based Sequential Quadratic Programming (SQP) optimization technique. The proposed GA-SQP-NN model for the ELMM is not just standard hybridization, it is for nonlinear linguistic learning processes when GA is used for global search and SQP for local refinement within the framework of a neural-network mapping structure. Another novelty is the use of error-based statistical measurements (TIC, MAD, RMSE) in the optimization cycle that creates convergence and precision than a GA-SQP. The results of the comparative measurements using the Lobatto method show that GA-SQP-NN is more reliable, more accurate, and more flexibly. There are also multi-run statistical verifications that prove the GA-SQP-NN model is stable. Therefore, the GA-SQP-NN structure is a usable and efficacious solver of complex mathematical models in the language acquisition area.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。