Quick Hidden Layer Size Tuning in ELM for Classification Problems
The extreme learning machine is a fast neural network with outstanding performance. However, the selection of an appropriate number of hidden nodes is time-consuming, because training must be run for several values, and this is undesirable for a real-time response. We propose to use moving average,...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Brno University of Technology
2024-06-01
|
| Series: | Mendel |
| Subjects: | |
| Online Access: | https://mendel-journal.org/index.php/mendel/article/view/299 |
| _version_ | 1827390707639255040 |
|---|---|
| author | Audi Albtoush Manuel Fernandez-Delgado Haitham Maarouf Asmaa Jameel Al Nawaiseh |
| author_facet | Audi Albtoush Manuel Fernandez-Delgado Haitham Maarouf Asmaa Jameel Al Nawaiseh |
| author_sort | Audi Albtoush |
| collection | DOAJ |
| description |
The extreme learning machine is a fast neural network with outstanding performance. However, the selection of an appropriate number of hidden nodes is time-consuming, because training must be run for several values, and this is undesirable for a real-time response. We propose to use moving average, exponential moving average, and divide-and-conquer strategies to reduce the number of training’s required to select this size. Compared with the original, constrained, mixed, sum, and random sum extreme learning machines, the proposed methods achieve a percentage of time reduction up to 98\% with equal or better generalization ability.
|
| first_indexed | 2024-03-08T16:56:39Z |
| format | Article |
| id | doaj.art-31345596ea104aa28f1569a53db7aa17 |
| institution | Directory Open Access Journal |
| issn | 1803-3814 2571-3701 |
| language | English |
| last_indexed | 2024-03-08T16:56:39Z |
| publishDate | 2024-06-01 |
| publisher | Brno University of Technology |
| record_format | Article |
| series | Mendel |
| spelling | doaj.art-31345596ea104aa28f1569a53db7aa172024-01-04T23:07:14ZengBrno University of TechnologyMendel1803-38142571-37012024-06-0130110.13164/mendel.2024.1.001Quick Hidden Layer Size Tuning in ELM for Classification ProblemsAudi Albtoush0Manuel Fernandez-Delgado1Haitham Maarouf2Asmaa Jameel Al Nawaiseh3Faculty of Computer Science and Information Technology, Jerash University, JordanSantiago De Compostela University, SpainSantiago De Compostela University, SpainSoftware Engineering Department, Mut'ah university, Jordan The extreme learning machine is a fast neural network with outstanding performance. However, the selection of an appropriate number of hidden nodes is time-consuming, because training must be run for several values, and this is undesirable for a real-time response. We propose to use moving average, exponential moving average, and divide-and-conquer strategies to reduce the number of training’s required to select this size. Compared with the original, constrained, mixed, sum, and random sum extreme learning machines, the proposed methods achieve a percentage of time reduction up to 98\% with equal or better generalization ability. https://mendel-journal.org/index.php/mendel/article/view/299Extreme Learning MachineNumber of Hidden NodesMoving AverageExponential Moving AverageDivide-and-conquer |
| spellingShingle | Audi Albtoush Manuel Fernandez-Delgado Haitham Maarouf Asmaa Jameel Al Nawaiseh Quick Hidden Layer Size Tuning in ELM for Classification Problems Mendel Extreme Learning Machine Number of Hidden Nodes Moving Average Exponential Moving Average Divide-and-conquer |
| title | Quick Hidden Layer Size Tuning in ELM for Classification Problems |
| title_full | Quick Hidden Layer Size Tuning in ELM for Classification Problems |
| title_fullStr | Quick Hidden Layer Size Tuning in ELM for Classification Problems |
| title_full_unstemmed | Quick Hidden Layer Size Tuning in ELM for Classification Problems |
| title_short | Quick Hidden Layer Size Tuning in ELM for Classification Problems |
| title_sort | quick hidden layer size tuning in elm for classification problems |
| topic | Extreme Learning Machine Number of Hidden Nodes Moving Average Exponential Moving Average Divide-and-conquer |
| url | https://mendel-journal.org/index.php/mendel/article/view/299 |
| work_keys_str_mv | AT audialbtoush quickhiddenlayersizetuninginelmforclassificationproblems AT manuelfernandezdelgado quickhiddenlayersizetuninginelmforclassificationproblems AT haithammaarouf quickhiddenlayersizetuninginelmforclassificationproblems AT asmaajameelalnawaiseh quickhiddenlayersizetuninginelmforclassificationproblems |