Applying fuzzy analytic hierarchy process to select models developed by abstraction and decision fusion architecture (case study: classification of Persian handwritten characters)
Purpose: This research presents an application-oriented approach for developing machine learning models that consider the trade-off between model accuracy, processing speed, and efficient resource utilization, focusing on applications such as wearable smart systems.Methodology: A set of models is de...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | fas |
| Published: |
Ayandegan Institute of Higher Education, Tonekabon,
2024-11-01
|
| Series: | تصمیم گیری و تحقیق در عملیات |
| Subjects: | |
| Online Access: | https://www.journal-dmor.ir/article_211082_f5e82d90e98bd8633aec8e7447c8e9d7.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832577845641084928 |
|---|---|
| author | Mohammad Safaei Somaye Moghari Mohammad Kazem Fallah Mehrdad Ghaznavi |
| author_facet | Mohammad Safaei Somaye Moghari Mohammad Kazem Fallah Mehrdad Ghaznavi |
| author_sort | Mohammad Safaei |
| collection | DOAJ |
| description | Purpose: This research presents an application-oriented approach for developing machine learning models that consider the trade-off between model accuracy, processing speed, and efficient resource utilization, focusing on applications such as wearable smart systems.Methodology: A set of models is developed based on the Abstraction and Decision Fusion Architecture (ADFA), and then, using a multi-criteria decision-making approach, the appropriate models for the intended application are identified. The proposed methodology has three main phases: 1) developing models based on the ADFA, 2) defining evaluation criteria, and 3) selecting models using the Fuzzy Analytic Hierarchy Process (FAHP).Findings: The experimental results of this research demonstrate the effectiveness of this approach in developing suitable machine learning models for applications related to wearable devices, such as smart glasses.Originality/Value: This research introduces three innovations: 1) the use of ADFA for developing models for the classification of Persian handwritten characters, 2) defining a new abstraction for summarizing handwritten character images, and 3) developing a fuzzy multi-criteria decision-making approach for mapping the developed models in the ADFA to real-world applications. |
| format | Article |
| id | doaj-art-e006f2bf8f2545e39d2cce8ee07af947 |
| institution | Kabale University |
| issn | 2538-5097 2676-6159 |
| language | fas |
| publishDate | 2024-11-01 |
| publisher | Ayandegan Institute of Higher Education, Tonekabon, |
| record_format | Article |
| series | تصمیم گیری و تحقیق در عملیات |
| spelling | doaj-art-e006f2bf8f2545e39d2cce8ee07af9472025-01-30T15:04:05ZfasAyandegan Institute of Higher Education, Tonekabon,تصمیم گیری و تحقیق در عملیات2538-50972676-61592024-11-019364966510.22105/dmor.2024.473217.1863211082Applying fuzzy analytic hierarchy process to select models developed by abstraction and decision fusion architecture (case study: classification of Persian handwritten characters)Mohammad Safaei0Somaye Moghari1Mohammad Kazem Fallah2Mehrdad Ghaznavi3Faculty of Mathematical Sciences, Shahrood University of Technology, Shahrood 3619995161, Iran.Faculty of Mathematical Sciences, Shahrood University of Technology, Shahrood 3619995161, Iran.Department of Computer Engineering, Chosun University, Gwangju 61453, Republic of Korea.Faculty of Mathematical Sciences, Shahrood University of Technology, Shahrood 3619995161, Iran.Purpose: This research presents an application-oriented approach for developing machine learning models that consider the trade-off between model accuracy, processing speed, and efficient resource utilization, focusing on applications such as wearable smart systems.Methodology: A set of models is developed based on the Abstraction and Decision Fusion Architecture (ADFA), and then, using a multi-criteria decision-making approach, the appropriate models for the intended application are identified. The proposed methodology has three main phases: 1) developing models based on the ADFA, 2) defining evaluation criteria, and 3) selecting models using the Fuzzy Analytic Hierarchy Process (FAHP).Findings: The experimental results of this research demonstrate the effectiveness of this approach in developing suitable machine learning models for applications related to wearable devices, such as smart glasses.Originality/Value: This research introduces three innovations: 1) the use of ADFA for developing models for the classification of Persian handwritten characters, 2) defining a new abstraction for summarizing handwritten character images, and 3) developing a fuzzy multi-criteria decision-making approach for mapping the developed models in the ADFA to real-world applications.https://www.journal-dmor.ir/article_211082_f5e82d90e98bd8633aec8e7447c8e9d7.pdfabstraction and decision fusion architectureclassification of persian handwritten charactersfuzzy analytic hierarchy processmachine learning |
| spellingShingle | Mohammad Safaei Somaye Moghari Mohammad Kazem Fallah Mehrdad Ghaznavi Applying fuzzy analytic hierarchy process to select models developed by abstraction and decision fusion architecture (case study: classification of Persian handwritten characters) تصمیم گیری و تحقیق در عملیات abstraction and decision fusion architecture classification of persian handwritten characters fuzzy analytic hierarchy process machine learning |
| title | Applying fuzzy analytic hierarchy process to select models developed by abstraction and decision fusion architecture (case study: classification of Persian handwritten characters) |
| title_full | Applying fuzzy analytic hierarchy process to select models developed by abstraction and decision fusion architecture (case study: classification of Persian handwritten characters) |
| title_fullStr | Applying fuzzy analytic hierarchy process to select models developed by abstraction and decision fusion architecture (case study: classification of Persian handwritten characters) |
| title_full_unstemmed | Applying fuzzy analytic hierarchy process to select models developed by abstraction and decision fusion architecture (case study: classification of Persian handwritten characters) |
| title_short | Applying fuzzy analytic hierarchy process to select models developed by abstraction and decision fusion architecture (case study: classification of Persian handwritten characters) |
| title_sort | applying fuzzy analytic hierarchy process to select models developed by abstraction and decision fusion architecture case study classification of persian handwritten characters |
| topic | abstraction and decision fusion architecture classification of persian handwritten characters fuzzy analytic hierarchy process machine learning |
| url | https://www.journal-dmor.ir/article_211082_f5e82d90e98bd8633aec8e7447c8e9d7.pdf |
| work_keys_str_mv | AT mohammadsafaei applyingfuzzyanalytichierarchyprocesstoselectmodelsdevelopedbyabstractionanddecisionfusionarchitecturecasestudyclassificationofpersianhandwrittencharacters AT somayemoghari applyingfuzzyanalytichierarchyprocesstoselectmodelsdevelopedbyabstractionanddecisionfusionarchitecturecasestudyclassificationofpersianhandwrittencharacters AT mohammadkazemfallah applyingfuzzyanalytichierarchyprocesstoselectmodelsdevelopedbyabstractionanddecisionfusionarchitecturecasestudyclassificationofpersianhandwrittencharacters AT mehrdadghaznavi applyingfuzzyanalytichierarchyprocesstoselectmodelsdevelopedbyabstractionanddecisionfusionarchitecturecasestudyclassificationofpersianhandwrittencharacters |