Less is more approach in optimization: a road to artificial intelligence

被引:3
|
作者
Mladenovic, Nenad [1 ,2 ]
Pei, Jun [2 ]
Pardalos, Panos M. [3 ,4 ]
Urosevic, Dragan [5 ]
机构
[1] Khalifa Univ, Dept Ind & Syst Engn, Abu Dhabi, U Arab Emirates
[2] Hefei Univ Technol, Sch Management, Hefei, Peoples R China
[3] Univ Florida, Gainesville, FL USA
[4] Higher Sch Econ, LATNA, Moscow, Russia
[5] Serbian Acad Arts & Sci, Math Inst, Belgrade, Serbia
关键词
Optimization; Less is more approach; Variable neighborhood search; Metaheuristics; Artificial intelligence; VARIABLE NEIGHBORHOOD SEARCH; FORMULATION;
D O I
10.1007/s11590-021-01818-w
中图分类号
C93 [管理学]; O22 [运筹学];
学科分类号
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要
The main idea of Less is more approach (LIMA) is using as fewer as possible ingredients to provide the best possible outcome. This approach has been used successfully almost in all the scientific and art disciplines. Recently, the idea has also been successfully explored in solving hard optimization problems. In this note we first define the dominance relation between two algorithms that includes their simplicity as well. Then we propose the general LIMA algorithm and discuss automatic ways to include common ingredients of all search algorithms, increasing the algorithms complexity in a systematic way. That kind of approach may represent a road from Optimization to Artificial Intelligence and Machine learning. Finally, we illustrate LIMA algorithm on two optimization problems and show its efficiency.
引用
收藏
页码:409 / 420
页数:12
相关论文
共 50 条