Extreme learning machine with a deterministic assignment of hidden weights in two parallel layers

详细信息    查看全文

• 作者：Pablo A. Henrí ; quez^a ; ^{pabhenriquez@alumnos.uai.cl}Author Vitae ; Gonzalo A. Ruz^a ; ^b ; ^{gonzalo.ruz@uai.cl}Author Vitae
• 关键词：Extreme learning machine ; Low-discrepancy points ; Parallel layers ; Regression ; Classification
• 刊名：Neurocomputing
• 出版年：2017
• 出版时间：22 February 2017
• 年：2017
• 卷：226
• 期：Complete
• 页码：109-116
• 全文大小：1567 K
• 卷排序：226

文摘

A new extreme learning machine architecture with two parallel layers is proposed. We use a deterministic assignment of the hidden weights using low-discrepancy points. The proposed PL-ELM algorithm is formulated for regression and classification. Overall the results showed that PL-ELM outperformed the traditional ELM.