Endüstri Mühendisliği Bölümü Koleksiyonu

Bu koleksiyon için kalıcı URI

https://hdl.handle.net/20.500.14034/37

Listeleniyor 1 - 20 / 93

Quasi-dynamic opposite learning enhanced Runge-Kutta optimizer for solving complex optimization problems
(Springer Heidelberg, 2024) Turgut, Oğuz Emrah; Turgut, Mert Sinan
The Runge-Kutta Optimization (RUNGE) algorithm is a recently proposed metaphor-free metaheuristic optimizer borrowing practical mathematical foundations of the famous Runge-Kutta differential equation solver. Despite its relatively new emergence, this algorithm has several applications in various branches of scientific fields. However, there is still much room for improvement as it suffers from premature convergence resulting from inefficient search space exploration. To overcome this algorithmic drawback, this research study proposes a brand-new quasi-dynamic opposition-based learning (QDOPP) mechanism to be implemented in a standard Runge-Kutta optimizer to eliminate the local minimum points over the search space. Enhancing the asymmetric search hyperspace by taking advantage of various positions of the current solution within the domain is the critical novelty to enrich general diversity in the population, significantly improving the algorithm's overall exploration capability. To validate the effectivity of the proposed RUNGE-QDOPP method, thirty-four multidimensional optimization benchmark problems comprised of unimodal and multimodal test functions with various dimensionalities have been solved, and the corresponding results are compared against the predictions obtained from the other opposition-based learning variants as well as some state-of-art literature optimizers. Furthermore, six constrained engineering design problems with different functional characteristics have been solved, and the respective results are benchmarked against those obtained for the well-known optimizers. Comparison of the solution outcomes with literature optimizers for constrained and unconstrained test problems reveals that the proposed QDOPP has significant advantages over its counterparts regarding solution accuracy and efficiency.
The wildfire suppression problem with multiple types of resources
(Elsevier, 2024) Avci, Mualla Gonca; Avcı, Mustafa; Battarra, Maria; Erdogan, Gunes
The frequency and impact of wildfires have considerably increased in the past decade, due to the extreme weather conditions as well as the increased population density. The aim of this study is to introduce, model, and solve a wildfire suppression problem that involves multiple types of fire suppression resources and their operational characteristics. Two integer programming (IP) formulations, a basic IP and its reformulation with combinatorial Benders' cuts, are presented. The performances of the proposed formulations are evaluated on a set of randomly generated instances. The results indicate that the proposed formulations are able to obtain high quality upper and lower bounds. Extensive numerical experiments are performed to analyse the effects of several operational constraints on the computational performance of the models. A case study arising in Yata & gbreve;an district of Mu & gbreve;la province of T & uuml;rkiye is presented.
Building sustainable resilient supply chain in retail sector under disruption
(Elsevier Sci Ltd, 2024) Ekinci, Esra; Sezer, Muruvvet Deniz; Mangla, Sachin Kumar; Kazancoglu, Yigit
Blockchain technologies have played a crucial role in transforming the retail industry, leading to remarkable advancements in recent times. Its pivotal role in managing risky environments by offering preventive and proactive measures cannot be overstated. The research contribution lies in introducing a set of criteria for assessing the adoption of Blockchain technology, specifically designed to evaluate the resilience of the retail sector. This study aims to ensure the establishment of a sustainable, resilient supply chain across diverse retail categories, particularly in the face of uncertain circumstances. A hybrid decision-making approach that combines the BestWorst Method (BWM) and Fuzzy TODIM has been employed to achieve these objectives. This study encompasses various types of retail companies to assess and compare their resilience levels by adopting Blockchain technology. The results of this study robustly suggest that speciality retailers with well-established, long-term partnerships are more predisposed to embrace and leverage the capabilities of Blockchain technologies. Conversely, discount retailers in Turkey face various challenges that impede their effective integration of Blockchain technologies. These challenges encompass collaborating with suppliers on short-term agreements and the unavailability of product tracking, among other factors. As a result, the outcomes of this study offer valuable insights for retailers in the sector, suggesting that they should consider modifying their operational strategies to better align with the adoption and integration of Blockchain technologies in the future.
Chaotic aquila optimization algorithm for solving phase equilibrium problems and parameter estimation of semi-empirical models
(Springer Singapore Pte Ltd, 2024) Turgut, Oğuz Emrah; Turgut, Mert Sinan; Kirtepe, Erhan
This research study aims to enhance the optimization performance of a newly emerged Aquila Optimization algorithm by incorporating chaotic sequences rather than using uniformly generated Gaussian random numbers. This work employs 25 different chaotic maps under the framework of Aquila Optimizer. It considers the ten best chaotic variants for performance evaluation on multidimensional test functions composed of unimodal and multimodal problems, which have yet to be studied in past literature works. It was found that Ikeda chaotic map enhanced Aquila Optimization algorithm yields the best predictions and becomes the leading method in most of the cases. To test the effectivity of this chaotic variant on real-world optimization problems, it is employed on two constrained engineering design problems, and its effectiveness has been verified. Finally, phase equilibrium and semi-empirical parameter estimation problems have been solved by the proposed method, and respective solutions have been compared with those obtained from state-of-art optimizers. It is observed that CH01 can successfully cope with the restrictive nonlinearities and nonconvexities of parameter estimation and phase equilibrium problems, showing the capabilities of yielding minimum prediction error values of no more than 0.05 compared to the remaining algorithms utilized in the performance benchmarking process.
Chaotic gradient based optimizer for solving multidimensional unconstrained and constrained optimization problems
(Springer Heidelberg, 2023) Turgut, Oğuz Emrah; Turgut, Mert Sinan
Gradient-based optimizer (GRAD) belongs to the recently developed population-based metaheuristic algorithms inspired by the development of Newton-type methods. Despite its new emergence, there are many successful applications of this optimizer in the existing literature; however, chaos integrated version of this algorithm has not been extensively studied yet. In his study, twenty-one different chaotic maps have been incorporated into the standard GRAD algorithm to maintain a reliable balance between exploration and exploitation mechanisms, which is not robustly constructed within the original algorithm. First ninety-nine thirty dimensional artificially generated optimization benchmark problems comprised of sixty-eight multimodal and thirty-one unimodal functions have been solved by these chaotic variants of the GRAD algorithm to determine the five best performing methods between them. Clear dominancy of the chaotic algorithms is clearly observed over the entire range of benchmark cases in terms of solution accuracy and robustness. Then, to validate the optimization capability of the chaos integrated GRAD algorithms, the best method among them is tested on fourteen constrained real world engineering problems, and its respective feasible results are benchmarked against those obtained from cutting edge metaheuristic optimizer. It is seen that the chaotic GRAD algorithm is able to effectively compete with other state-of-art algorithms on both solving unconstrained and constrained engineering problems. Moreover, it is observed that the Chebyshev chaotic map improved GRAD algorithm outperforms its contemporaries in both unconstrained and constrained cases.
İki amaçlı tesis yerleşimi problemi için bir yinelemeli yerel arama algoritması
(2021) Avcı, Mustafa
Tesis yerleşimi, modern üretim sistemlerinde karşılaşılan en önemli sorunlardan biridir. Bu çalışmada hem nicel hem de nitel hedeflerin birleştirildiği iki amaçlı bir tesis yerleşimiproblemi(İA-TYP) ele alınmaktadır. Burada nicel amaç toplam malzeme taşıma maliyetinin en aza indirgenmesidir. Nitel amacımız ise toplam yakınlık derecelendirme puanlarının maksimize edilmesidir. Problemi çözmek için Değişken Komşuluklu İniş (DKİ) algoritmasının bir saptırma mekanizması ile birleştirildiği bir Yinelemeli Yerel Arama (YYA) algoritması önerilmiştir. Önerilen algoritmanın performansı önceki çalışmalarda sunulan çözüm algoritmalarının elde ettikleri çözümler baz alınarakdeğerlendirilmiştir. Hesaplama sonuçları, önerilen algoritmanın İA-TYP örneklerine yüksek kaliteli çözümler üretebildiğini göstermektedir.
A comparative analysis of constraint-handling mechanisms for solving engineering design problems
(2021) Gölcük, İlker
Optimization problems have numerous real-life applications in science and engineering. The engineering design problems are usually subject to various constraints. Although many state-of-the-art metaheuristic optimization algorithms have been developed during the last decades, these algorithms require additional constraint-handling mechanisms to cope with constrained optimization problems. Therefore, selecting a suitable constraint-handling mechanism requires extensive trial-and-error experiments, which is time-consuming and demanding. In this study, a comparative analysis of the eight constraint handling mechanisms is carried out, guiding decision-makers in their optimization practices. The constraint-handling techniques are used along with the Whale Optimization Algorithm (WOA), and 19 real-life mechanical design problems, which are also part of the CEC2020 benchmark suite, are tested in the experimental analysis. The nonparametric statistical analysis incorporating Nemenyi and Holm post-hoc procedures shows that the inverse tangent constraint-handling and eclectic penalty methods exhibit high performance in real-life mechanical design problems.
Categorization of customer complaints in food industry using machine learning approaches
(2022) Kılınç, Deniz; Doğan, Onur; Bozyiğit, Fatma
Customer feedback is one of the most critical parameters that determine the market dynamics of product development. In this direction, analyzing product-related complaints helps sellers to identify the quality characteristics and consumer focus. There have been many studies conducted on the design of Machine Learning (ML) systems to address the causes of customer dissatisfaction. However, most of the research has been particularly performed on English. This paper contributes to developing an accurate categorization of customer complaints about package food products, written in Turkish. Accordingly, various ML algorithms using TF-IDF and word2vec feature representation strategies were performed to determine the category of complaints. Corresponding results of Linear Regression (LR), Naive Bayes (NB), k Nearest Neighbour (kNN), Support Vector Machine (SVM), Random Forest (RF), and Extreme Gradient Boosting (XGBoost) classifiers were provided in related sections. Experimental results show that the best-performing method is XGBoost with TF-IDF weighting scheme and it achieves %86 F-measure score. The other considerable point is word2vec based ML classifiers show poor performance in terms of F-measure compared to the TF-IDF term weighting scheme. It is also observed that each experimented TF-IDF based ML algorithm gives a more successful prediction performance on the optimal subsets of features selected by the Chi Square (CH2) method. Performing CH2 on TF-IDF features increases the F-measure score from 86% to 88% in XGBoost.
Process mining methodology for digital processes under smart campus concept
(2022) Cengiz, Esra; Doğan, Onur
Digital transformation affects universities as well as many industries. Universities are increasingly using various digital resources and systems to manage their knowledge. The smart campus, on the other hand, supports informed decision-making by integrating these resources and systems. Process mining provides real insights for digital transformation, allowing processes to be examined more transparently. This study aims to examine the proposed project implementation processes related to the smart university with the process mining methodology. For this purpose, 32 completed projects submitted to İzmir Bakırçay University Scientific Research Projects Coordinatorship (BAPK) with the proposed methodology adapted from Deming's continuous improvement cycle were examined. The data are taken from two different pages in the project automation system. According to the research findings, Projects are grouped into three categories: Guided Projects (GDM, 5 projects), Graduate Thesis Projects (TEZ, 5 projects), and Career Start Support Projects (KBP, 22 projects). 40.6% (13 projects) of the applications went directly to the project review stage, while 19 (59.4%) needed procedural correction. Considering the time from the creation of the application of 32 projects to the signing of the contract, it is seen that the arithmetic average of the cycle time is 15.1 weeks, and the median average is 52.5 days. The notable difference between arithmetic and median mean is that very few projects are of long duration. Procedural adjustments affect project evaluation cycle time by an additional 14 days. The carelessness or lack of knowledge of the applicants extends the cycle time of the process from 15 days to 53 days. The total duration of unnecessary waiting time in the process is 17 days. This study primarily proposes that non-digital processes should be digitized as soon as possible.
Müşteri memnuniyetinin süreç odaklı değerlendirilmesi: Bir çağrı merkezinde süreç madenciliği uygulaması
(2021) Doğan, Onur; Ayyar, Başak; Çağıl, Gültekin
Firmaların varlıklarını sürdürebilmeleri için müşteri odaklı bir yaklaşım sergilemeleri büyük avantaj sağlamaktadır. Müşteriler artık kendisine sunulanı değil, beklentilerini karşılayan ürün veya hizmeti satın almaktadır. Bu sebepten dolayı firmalar müşterileri dinlemek, onların ihtiyaç ve beklentilerini karşılamak amacıyla çeşitli stratejiler izlemektedirler. Bu stratejik birimlerden birisi de firma ile müşteri arasında iletişimi sağlayan çağrı merkezleridir. Firmalar çağrı merkezleri ile müşterilerin sorunlarını çözerek memnuniyet düzeyini arttırmaya çalışırlar. Bununla birlikte, müşteri memnuniyetinin yönetilebilir, ölçülebilir ve karşılaştırılabilir olması için memnuniyet anketi uygulamaktadırlar. Bu çalışmanın amacı, müşteri memnuniyet anketlerinin de göz önüne alındığı süreç madenciliği yardımıyla müşteri memnuniyetinin değerlendirilmesidir. Firmanın IT sisteminde tutulan çağrı merkezi bölümüne ait 11 günlük (18 – 28 Mart 2019) tarihleri arasındaki 42185 olay günlüğü (event log) süreç madenciliği için kullanılmıştır. Öncelikle memnuniyet anketine katılan ve katılmayan müşterilerin sistem içindeki akışı çıkarılmıştır. Ardından ankete katılan müşteriler arasında 5 (çok memnunum) puan verenlerin ve diğer müşterilerin süreçleri incelenmiştir. Çalışmada ulaşılan sonuçlardan biri, çağrı sırasında bekletilen müşterilerin daha yüksek puan vermesidir. Bu sonuç müşterilerin bekletilmekten ziyade ihtiyacının karşılanmasına daha çok önem verdiğini göstermektedir. Bu açıdan çalışma, firmanın süreç odaklı yönetimine, süreç iyileştirme çalışmalarına, strateji çalışmalarına yol gösterebilecek özelliktedir.
Enterprise resource planning selection using fuzzy entropy-based fuzzy MOORA method: case study in a bearing company
(2022) Durmuş, Beyzanur; Akpınar, Muhammet Enes
Seçim problemleri işletmeler açısından sıklıkla karşılaşılan ve karar vermesi zor olan problem tiplerindendir. Zor problem olmasının sebebi birçok kriter ve alternatifin aynı anda dikkate alınması gerektiği içindir. Bu problemlerin çözümü için genellikle çok kriterli karar verme yaklaşımları kullanılmaktadır. Seçim problemleri hayatın her aşamasında karşılaştığı için çok fazla çeşitlilik gösterebilmektedir. Bu çal ışmada bir işletmenin kurumsal kaynak planlaması (KKP) seçim süreci ele al ınmıştır. Yeni bir yazılım satın almak isteyen i şletmenin satın alma departmanı birçok kriter ve alternatif yazılım belirlemiştir. Bu kriterlerin en uygun düzeyde karşılandığı alternatif yazılımın seçilmesi planlanmıştır. Bu problemin çözümü için kriter ağırlıkların belirlenmesi aşamasında bulanık Entropi yöntemi kullanılmıştır. Yazılım alternatiflerinin değerlendirilmesi sürecinde bulanık Oran Analiziyle Çok Amaçlı Optimizasyon (MOORA) yöntemi kullanılmış ve yazılımlardan en uygun olanına karar verilmiştir. Çalışma sonucunda belirlenen üç yazılım sisteminden en uygun olanın üçüncü yazılım sistemi olduğu görülmüştür.
The effect of brand-relationship quality on positive word-of-mouth intention
(IEOM Society, 2021) Akın Mahmut Selami; Doğan Onur
Measurement and management of consumer relationships with brands are critical issue that marketers have focused on for years. The quality of the relationship between the consumer and brand affects behavioral intention in favor of the brand. This research aims to reveal the effect of consumer-brand relationship quality (BRQ) on positive word-of-mouth (WoM) intention. The BRQ components were applied on the sample of private university students in Istanbul, Turkey. The questionnaire form with 5-point Likert scale was distributed and 199 valid surveys were obtained. Cronbach's Alfa was examined to internal consistency of the scales, and exploratory factor analysis was conducted to assess constructive validity. Multiple regression analysis resulted in love/passion, commitment, and self-connection, which are the dimensions of BRQ, have a positive WoM intention. On the other hand, the intimacy factor has no significant effect. © IEOM Society International.
Applications of machine learning approaches to combat COVID-19: A survey
(Elsevier, 2022) Tiwari Sanju Mishra; Doğan Onur; Jabbar M. Akhil; Shandilya Shishir K; Ortiz-Rodríguez Fernando; Bajpai Sailesh; Banerjee Sourav
Machine learning (ML) and artificial intelligence (AI) approaches are prominent and well established in the field of health-care informatics. Because they have a more productive ability to predict, they are successfully applied in several health-care applications. ML approaches are needed thanks to the unsatisfactory experience of the novel virus, considerable ambiguity, complicated social circumstances, and inadequate accessible data. Several approaches have been applied as a tool to combat and protect against the new diseases. The COVID-19 outbreak has rapid growth, so it is not easy to predict the patients and resources within a specified time. ML is a strong approach in the fighting against the pandemic such as COVID-19. It is found significant to predict the susceptible, infected, recovered, or exposed persons and can assist the control strategies to block the spread of infections. This study critically examines the appropriateness and contribution of AI/ML methods on COVID-19 datasets, enhancing the understanding to apply these methods for quick analysis and verification of pandemic databases. © 2022 Elsevier Inc. All rights reserved.
Machine learning applications for Fraud Detection in finance sector
(Springer, 2022) Taşer, Pelin Yıldırım; Bozyiğit, Fatma
Due to advances in information technology, instantaneous accessibility to financial services through digital channels has increased. Although digital platforms’ usage makes an individual’s life more comfortable, it may also cause some critical consequences like financial fraud which causes critical losses for companies in the industrial sector, investors, and governments. Identification of frauds can be challenging task for a human because it may be necessary to analyse high volume data during long time periods. An alternative is to use financial data as a fraud detection tool to automatically classify fraudulent activities. Currently, there are many practical solutions for automatically detect frauds in the finance domain. In this chapter, we examined on three different fraud types (bank fraud, insurance fraud, and corporate fraud) in finance sector and reviewed the studies using machine learning methods to detect financial fraud in a detailed manner. The findings from this review show that most commonly applied algorithms for financial fraud detection are Decision Tree, Support Vector Machine (SVM), Artificial Neural Networks (ANN), and Random Forest and most of machine learning-based studies were performed in bank fraud field. This chapter also reveals that deep learning and ensemble-based machine learning applications has been frequently preferred in recent years to improve detection performance of the frauds in finance sector. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
Prioritizing occupational safety risks with fuzzy FUCOM and fuzzy graph theory-matrix approach
(Gazi Univ, Fac Engineering Architecture, 2023) Gölcük, İlker; Durmaz, Esra Duygu; Şahin, Ramazan
In this study, a new failure mode and effects analysis (FMEA) model is proposed for evaluating occupational safety risks. In the classical FMEA, risk priority numbers (RPNs) are calculated by multiplying the risk scores of the occurrence, severity, and detectability. However, RPN numbers generated by classical FMEA have been the subject of severe criticism in the literature. To overcome the drawbacks of the classical FMEA, this study proposes a new Multiple Attribute Decision Making (MADM) model. The proposed risk evaluation model combines the full consistency method (FUCOM) and graph theory-matrix approach (GTMA) under a fuzzy environment. The risk scores of failure models and the weights of risk factors have been obtained using the fuzzy FUCOM method. On the other hand, the RPN value of each failure mode is calculated by utilizing fuzzy GTMA. Fuzzy GTMA considers all possible dependencies among risk factors, which in turn produces more accurate rankings. The fuzzy judgements of the decision makers are aggregated by using the least squares distance method. The proposed model is implemented in a real-life case study and the failure modes are ranked.
Multi-liquid repellent, fluorine-free, heat stable SLIPS via layer-by-layer assembly
(Elsevier, 2022) Kasapgil, Esra; Erbil, H. Yıldırım; Şakir, İlke Anaç
Slippery liquid-infused porous surfaces (SLIPS) are an emerging class of bio-inspired materials attracting re-searchers due to their good anti-fouling, anti-icing, and self-cleaning performance. The fluorinated lubricants which are used in the preparation of multi-liquid repellent SLIPS are not suitable for practical applications because of the lubricant losses by evaporation and water drop cloaking and also they cause environmental and health problems. In this study, layer-by-layer (LbL) assembly technique via a capping approach is used to obtain PDMS-based SLIPSs. Amino terminated PDMS was used as a homofunctional polymer layer and three different co-polymers of anhydride [poly(ethylene-alt-maleic anhydride) PEMA, poly(isobutylene-alt-maleic anhydride) PIMA and poly(maleic anhydride-alt-1-octadecene) PMAO] were used as the capping layers to give LbL layers having advancing contact angles between 115 and 120? with the change of the type of anhydride copolymer used and the constituent concentration. Silicone oil was used as the lubricant giving a fluorine-free LbL-SLIPS having a contact angle hysteresis as low as 5-6 repelling water and various other liquids such as glycerol, ketchup and soy sauce by preserving the low contact angle hysteresis for long durations. Moreover, LbL-SLIPS samples also had good heat resistance (to 80 ?C for 8 days) and chemical stability against acids and bases and were durable under the flow of continuous water droplets for a period of 4 h without any damage to SLIPS properties. The best performing LbL-SLIPS sample was prepared using the PEMA capping layer where the decrease in silicone oil thickness and in-crease in contact angle hystresis were found to be lowest under continuous water droplet flow after 4 h.
Interval type-2 fuzzy development of FUCOM and activity relationship charts along with MARCOS for facilities layout evaluation
(Elsevier, 2022) Gölcük, İlker; Durmaz, Esra Duygu; Şahin, Ramazan
This paper presents a new interval type-2 fuzzy (IT2F) multiple attribute decision making (MADM) model for evaluating facility layout alternatives. To this end, The Full Consistency Method (FUCOM) is extended to IT2F sets, and the mathematical model of IT2F-FUCOM is proposed. IT2F-FUCOM method requires a manageable number of pairwise comparisons and can deal with imprecision and uncertainty. Furthermore, the Activity Relationship Charts (ARCs), which are powerful instruments to articulate preferences regarding closeness between departments in the facility design, are also modeled through IT2F sets. The proposed IT2F-ARCs are incorporated into the decision hierarchy as a cost-type criterion, a novel integration between ARCs and the MADM is provided. Finally, the state-of-the-art IT2F MADM method of Measurement Alternatives and Ranking according to the Compromise Solution (MARCOS) is utilized to rank the alternatives. A real-life case study is conducted in order to demonstrate the applicability of the proposed model. (C) 2022 Elsevier B.V. All rights reserved.
Impact of epidemic outbreaks (COVID-19) on global supply chains: A case of trade between Turkey and China
(Elsevier Science Inc, 2023) Kazançogğu, Yiğit; Ekinci, Esra; Mangla, Sachin Kumar; Sezer, Mürüvvet Deniz; Özbiltekin-Pala, Melisa
COVID-19 has negative impacts on supply chain operations between countries. The novelty of the study is to evaluate the sectoral effects of COVID-19 on global supply chains in the example of Turkey and China, considering detailed parameters, thanks to the developed System Dynamics (SD) model. During COVID-19 spread, most of the countries decided long period of lockdowns which impacted the production and supply chains. This had also caused decrease in capacity utilizations and industrial productions in many countries which resulted with imbalance of maritime trade between countries that increased the freight costs. In this study, cause and effect relations of trade parameters, supply chain parameters, demographic data and logistics data on disruptions of global supply chains have been depicted for specifically Turkey and China since China is the biggest importer of Turkey. Due to this disruption, mainly exports from Turkey to China has been impacted in food, chemical and mining sectors. This study is helpful to plan in which sectors; the actions should be taken by the government bodies or managers. Based on findings of this study, new policies such as onshore activities should consider to overcome the logistics and supply chain disruptions in global supply chains. This study has been presented beneficial implications for the government, policymakers and academia.
Fuzzy association rule mining approach to identify e-commerce product association considering sales amount
(Springer Heidelberg, 2022) Doğan, Onur; Kem, Furkan Can; Öztayşi, Başar
Online stores assist customers in buying the desired products online. Great competition in the e-commerce sector necessitates technology development. Many e-commerce systems not only present products but also offer similar products to increase online customer interest. Due to high product variety, analyzing products sold together similar to a recommendation system is a must. This study methodologically improves the traditional association rule mining (ARM) method by adding fuzzy set theory. Besides, it extends the ARM by considering not only items sold but also sales amounts. Fuzzy association rule mining (FARM) with the Apriori algorithm can catch the customers' choice from historical transaction data. It discovers fuzzy association rules from an e-commerce company to display similar products to customers according to their needs in amount. The experimental result shows that the proposed FARM approach produces much information about e-commerce sales for decision-makers. Furthermore, the FARM method eliminates some traditional rules considering their sales amount and can produce some rules different from ARM.
An improved arithmetic optimization algorithm for training feedforward neural networks under dynamic environments
(Elsevier, 2023) Gölcük, İlker; Özsoydan, Fehmi Burçin; Durmaz, Esra Duygu
This paper proposes an improved Arithmetic Optimization Algorithm (AOA) to train artificial neural networks (ANNs) under dynamic environments. Despite many successful applications of metaheuristic training of ANNs, these studies assume static environments, which might not be realistic in real-world nonstationary processes. In this study, the training of ANNs is modeled as a dynamic optimization problem, and the proposed AOA is used to optimize connection weights and biases of the ANN under the presence of concept drift. The proposed method is designed to work for classification tasks. The performance of the proposed algorithm has been tested on twelve dynamic classification problems. Comparative analysis with state-of-the-art metaheuristic optimization algorithms has been provided. The superiority of the compared algorithms has been verified using nonparametric statistical tests. The results show that the improved AOA outperforms compared algorithms in training ANNs under dynamic environments. The findings demonstrate the potential of improved AOA for dynamic data-driven applications.(c) 2023 Elsevier B.V. All rights reserved.

Güncel Gönderiler