International Journal of Advanced Natural Sciences and Engineering Researches
https://as-proceeding.com/index.php/ijanser
<p>International Journal of Advanced Natural Sciences and Engineering Researches (IJANSER) publishes regular research papers, reviews, letters, and communications covering all aspects of engineering and natural sciences. Our aim is to publish novel / improved methods/approaches of these field to benefit the community, open to everyone in need of them. There is no restriction on the length of the papers or colors used. The method/approach must be presented in detail so that the results can be reproduced.</p>Umut Özkayaen-USInternational Journal of Advanced Natural Sciences and Engineering Researches2980-0811Enhancing Customer Churn Prediction in the Finance Sector through Explainable AI and Machine Learning
https://as-proceeding.com/index.php/ijanser/article/view/2827
<p>Customer churn prediction has emerged as a critical research domain in various industries, <br>including telecommunications, retail, and finance, due to its significant impact on business profitability, <br>customer satisfaction, and long-term sustainability. Churn refers to the rate at which customers terminate <br>their relationship with a company, necessitating the development of accurate predictive models to <br>facilitate effective retention strategies. In this context, machine learning models have demonstrated <br>substantial potential by processing large datasets to identify patterns indicative of customer attrition. <br>However, despite their predictive accuracy, the widespread adoption of these models is often limited by <br>their lack of transparency, commonly referred to as the “black box” problem. This challenge is <br>particularly observed in sectors such as finance and risk management, where customer trust is of <br>paramount importance. Explainable AI (XAI) addresses this limitation by enhancing the interpretability <br>of machine learning models, enabling stakeholders to comprehend the rationale behind predictions while <br>maintaining model performance. This study investigates the integration of XAI methodologies into <br>customer churn prediction models within the financial sector, with a focus on Eminevim, addressing the <br>challenges posed by complex data and the necessity for actionable insights. The performance and <br>interpretability of various machine learning algorithms, such as Random Forest, XGBoost, Light-GBM, <br>and CatBoost are assessed utilizing explainability techniques such as SHAP (Shapley Additive <br>Explanations). The findings demonstrate that XAI-augmented churn prediction models not only preserve <br>high predictive accuracy but also enhance transparency, empowering financial institutions to make <br>informed, data-driven decisions that mitigate customer attrition and promote long-term business <br>sustainability.</p>Fidan KhalilbayliCengiz Sertkaya
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-0191018The Impact of Relationship Marketing in Financial Institutions in Emerging Markets: Evidence from Albania
https://as-proceeding.com/index.php/ijanser/article/view/2828
<p>In recent years, managers of financial institutions have increasingly acknowledged the strategic <br>importance of retaining existing customers as a critical determinant of competitiveness and long-term <br>sustainability. Marketing literature consistently highlights the advantages of cultivating a loyal customer base, <br>including reduced churn, enhanced profitability, and stronger advocacy. At the core of customer loyalty lies <br>the quality and continuity of the relationship between the client and the institution. The deeper and more <br>sustained this relationship, the greater the mutual value generated for both parties. <br>This study examines the role of relationship marketing in the context of financial institutions in Albania, an <br>emerging market where competition and customer expectations are intensifying. Specifically, it investigates <br>how Customer Relationship Management (CRM) practices influence customer satisfaction and behavioral <br>loyalty. A quantitative research design was employed, with data collected through structured questionnaires <br>administered to clients and employees of selected financial institutions across Albania. A purposive sampling <br>method ensured the inclusion of respondents with direct experience in banking and other financial services. A <br>total of over 400 valid responses (300 customers and 100 employees) were analyzed using SPSS version 25. <br>The Friedman test was applied to assess differences in customer awareness across CRM practices, while a <br>Neural Network Model was used to predict satisfaction based on demographic and behavioral variables. <br>Findings revealed that CRM attributes are not perceived equally, with “CRM enhances customer loyalty” <br>emerging as the most influential factor. The results underscore that effective CRM implementation in Albanian <br>financial institutions relies not only on technological infrastructure but also on human interaction, service <br>accessibility, and personalized engagement strategies. Hence, financial institutions in Albania are advised to <br>prioritize customer-centric CRM strategies, continuous staff training, and targeted communication to <br>strengthen trust, satisfaction, and long-term loyalty.</p> <p> </p>Dr. Ledia SulaDr. Bajram R. Hasani
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-01910920Ön Şarjlı Gerilim Kaynaklı Dönüştürücü
https://as-proceeding.com/index.php/ijanser/article/view/2829
<p>Güç elektroniği sistemleri, enerji dönüşümü ve kontrolü açısından büyük bir öneme sahiptir. Bu <br>sistemlerin verimliliği ve güvenliği, kullanılan dönüştürücüler ve kontrol tekniklerine doğrudan bağlıdır. <br>Gerilim kaynaklı dönüştürücüler, doğru akım gücünü alternatif akım gücüne dönüştürmek için yaygın <br>olarak kullanılan bir güç dönüştürücüsüdür. Özellikle, ön şarjlı gerilim kaynaklı dönüştürücüler, şebeke <br>bağlantılı sistemlerde veya batarya tabanlı güç sistemlerinde tercih edilir. Bu sistem, devreye giriş sırasında <br>oluşabilecek yüksek akım darbelerini engelleyerek ekipmanları korumak amacıyla tasarlanmıştır. Bu <br>çalışmada, ön şarjlı ve kapalı devre kontrollü 3 fazlı gerilim kaynaklı dönüştürücü devre yapısı Plexim <br>ortamında modellenmiş ve incelenmiştir.</p> <p> </p>Yasin BEKTAŞ
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-019102125Exploring the Educational Usability of Popular 3D Scanning Applications
https://as-proceeding.com/index.php/ijanser/article/view/2830
<p>This study explores the educational usability of four popular 3D scanning applications - LumaAI, <br>Polycam, KIRI Engine, and 3D Scanner App, by comparing their performance, user experience, and <br>potential integration into teaching and learning contexts. Three of the applications (LumaAI, Polycam, and <br>KIRI Engine) were tested on an Android device, while the 3D Scanner App was examined on an Apple <br>iPad Pro to leverage its built-in LiDAR sensor. To assess their ability to handle challenging conditions, the <br>target object was placed under a transparent glass dome, allowing us to evaluate how each application <br>processes light refraction and reflections during the scanning process. <br>The analysis focused on two main aspects: (1) the quality and accuracy of the generated 3D models, and <br>(2) the overall user experience, including accessibility, workflow simplicity, and platform compatibility. <br>These criteria were then evaluated in relation to their potential application in educational settings, <br>particularly for subjects where interactive visualization, digital modeling, or augmented learning <br>experiences can enhance engagement and understanding. <br>Findings indicate differences among the applications in terms of usability, rendering fidelity, and suitability <br>for classroom use. While some applications demonstrated robustness and ease of use, others struggled with <br>reflective surfaces or required more advanced features. The results highlight both opportunities and <br>limitations of current mobile 3D scanning tools, providing insights for educators seeking to integrate such <br>technologies into pedagogical practice.</p>Mgr. Gergely KocsisIng. Ondrej TakáčPaed Dr. Bence PásztorMgr. László Halász
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-019102637Hastane Yangınlarının Etkileri: Sağlık, Çevre ve Toplum Üzerine Bir İnceleme
https://as-proceeding.com/index.php/ijanser/article/view/2831
<p>Yangınlar, çevre ve sağlık sistemleri üzerinde önemli etkiler yaratabilen doğal ve insan kaynaklı <br>olaylardır. Özellikle hastane yangınları, sağlık hizmetlerinin sürekliliğini tehdit eden, altyapı ve hasta <br>güvenliğini doğrudan etkileyen olaylardır. Yangın, her yerde görülebilen, ısı kaynağı, yanıcı madde ve <br>yeterli seviyedeki oksijenin bir araya gelmesiyle oluşan bir felaket türüdür. Yangınlar orman yangınları gibi <br>doğada görülebildiği gibi binalarda da görülebilmektedir. Hastaneler insanlara sağlık hizmeti veren <br>binalardır. Ancak bu binalarda da yangınlara rastlanabilmektedir. Özellikle yoğun bakım ünitelerinin ve <br>yatan hastaların olduğu hastanelerdeki yangınlar yardıma muhtaç insanların hayatı söz konusu olduğundan <br>daha kritik bir hal almaktadır. Türkiye’de hastane yangınları konusunda geniş kapsamlı ve güvenilir bir <br>kaynak bulunmamaktadır. Ülke genelinde tutulan kayıtların toplandığı ulusal bir veri merkezi de maalesef <br>yoktur. Yangınlar ile ilgili veriler lokal ve dağınık olarak çeşitli resmi kurumlarda tutulmaktadır. Bu <br>çalışmayla ülkemizde 2013-2022 tarihleri arasında sosyal medyaya yansımış hastane yangınları kayıt altına <br>alınmıştır. Kayıt altına alınan hastane yangınları MS Excel ve Spss 22.0 ile frekans dağılımı <br>değerlendirilmiştir. Toplam 271 hastane yangını tespit edilmiştir. En önemli yangın nedeni olarak elektrik <br>tesisatı ve en çok yangın çıkan bölüm ise hasta servisi/poliklinikler olarak tespit edilmiştir. Bu çalışma, <br>çevre ve hastane yangınlarının sağlık, çevre ve toplum üzerindeki etkilerini incelemeyi amaçlamaktadır.</p>Mehmet ErensoySerden Başak
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-019103852Sustainability as a Driver of Digital Transformation: The Role of Emerging Technologies
https://as-proceeding.com/index.php/ijanser/article/view/2832
<p>Digital transformation has emerged as a defining force in reshaping economic, social, and <br>environmental domains, with sustainability increasingly influencing the trajectory of this transformation. <br>The relationship between digitalization and sustainability is mutually reinforcing: sustainability sets the <br>direction for the development, adoption, and diffusion of digital technologies, while digital tools and <br>processes provide essential mechanisms for realizing sustainable outcomes. Emerging technologies such as <br>artificial intelligence (AI), big data analytics, the Internet of Things (IoT), and blockchain are playing a <br>pivotal role in enabling greener business models, enhancing transparency, optimizing resource efficiency, <br>and reducing environmental footprints. At the same time, organizations face challenges related to high <br>implementation costs, data security concerns, regulatory complexity, and digital skill gaps. This research <br>examines the multidimensional impact of sustainability on digital transformation through both SWOT and <br>PESTLE analyses. The findings reveal that sustainability not only drives innovation and efficiency but also <br>creates new opportunities for circular economy models, global collaboration, and investment in green <br>technologies. However, threats such as digital divides, technological lock-ins, regulatory challenges and <br>the rising energy demand of digital infrastructures pose significant risks. By combining theoretical insights <br>with strategic analysis, this study underscores the importance of organizational capabilities, governance, <br>and innovation in leveraging digital transformation to achieve long-term sustainability goals. The <br>discussion contributes to the growing body of literature by offering a comprehensive framework to <br>understand how sustainability can shape digital transformation in practice and by highlighting the need for <br>holistic approaches to address both opportunities and challenges, including political, economic, social, <br>technological, legal, and environmental factors.</p>Ayşenur ERDİL
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-019105360Derin Öğrenme Tabanlı Melanoma Tespiti: Karşılaştırmalı Model Analizi
https://as-proceeding.com/index.php/ijanser/article/view/2833
<p>Cilt kanseri, cilt hücrelerinin kontrolsüz bir biçimde büyümesi ile meydana gelen bir kanser <br>biçimidir. Melanom, bu kanser türünün en nadir görülen fakat en çok ölüme sebep olan türüdür. Bu <br>çalışmada, melanoma tespiti amacıyla dermoskopi görüntülerini kullanarak bir sınıflandırma modeli <br>geliştirilmiştir. Benign (iyi huylu) ve malign (kötü huylu) olarak etiketlenmiş 10.000 deri üzeri iz <br>fotoğrafından oluşan bir veri seti kullanılarak, deri üzerinde bulunan izlerin kötü huylu melanoma cilt <br>kanseri mi yoksa iyi huylu mu olduğunu bulma amacıyla bir sınıflandırma modeli geliştirilmiştir. <br>Bu çalışmada, üç farklı derin öğrenme mimarisi (AlexNet, EfficientNet ve ResNet50) kullanılarak <br>melanoma tespiti için karşılaştırmalı bir analiz gerçekleştirilmiştir. Modellerin genelleme yeteneğinin <br>artması, doğruluk oranının yükselmesi ve aşırı öğrenmeyi engellemek için çeşitli veri artırma yöntemleri <br>ve görüntü zenginleştirme yöntemleri kullanılmıştır. Bu yöntemler modellerin farklı görüntülerle bile daha <br>etkili performans göstermesini sağlamıştır. Bu çalışmada elde edilen Out-of-Fold (OOF) ROC-AUC <br>skorları 0.952(AlexNet), 0.973 (EfficientNet) ve 0.963 (ResNet50) olarak hesaplanmıştır. Bu değerler, <br>modellerin genel performansını ve genelleme yeteneğini doğrulamaktadır. Modellerin doğrulama veri seti <br>üzerindeki doğruluk oranı maksimum %97.3 olarak ölçülmüş ve bu da modellerin yüksek bir doğruluk <br>oranına ulaştığını göstermektedir. <br>Sonuç olarak, bu araştırma, yapay zekâ ile kanser teşhisi gibi kritik alanlardaki başarısını göstermektedir. <br>Çalışmanın temel başarı unsurlarından biri olarak; transfer öğrenme yöntemi ile büyük veri setlerinde <br>eğitilmiş modellerin daha küçük ölçekli medikal veri kümelerine başarıyla uyum sağlayabilmesi olarak <br>belirtilebilir. Bu güncel literatür karşılaştırması, çalışmamızın mevcut en iyi uygulamalarla rekabet edebilir <br>ve bazı durumlarda daha iyi sonuçlar elde edebildiğini göstermektedir. Özellikle EfficientNet modelimizin <br>performansı, literatürdeki benzer çalışmaların literatürdeki bazı çalışmaların önüne geçmiştir.</p>Halil İbrahim KOCAGÖZDoc. Dr. Ömer KASIM
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-019106168Machine Learning in Biosciences: A Review of Applications
https://as-proceeding.com/index.php/ijanser/article/view/2835
<p>Studies have shown that only about 2% of the genome encodes proteins, while the remaining 98% consists of non-coding RNAs (ncRNAs). Based on length, ncRNAs are classified as small (<200 nt) or long (>200 nt) and play key roles in biological processes. Experimentally verified associations between ncRNAs (miRNAs, lncRNAs, circRNAs) and diseases remain limited, since laboratory studies are costly and time-consuming. Thus, computational approaches have become essential for predicting disease related ncRNAs. Similarly, drug-target interactions are vital for drug discovery, as drugs act by binding to and inhibiting target molecules. Yet, experimental identification of these interactions is expensive, driving the development of computational prediction methods. Microbes also influence human health, with microbiomes playing essential physiological roles. Identifying disease-related microbes is crucial, but experimental approaches are limited by cost and time. Hence, computational methods are widely employed. Overall, computational strategies can be grouped into score functions, network-based algorithms, multi source biological integration, and machine learning. This review highlights machine learning approaches for predicting ncRNA-disease associations, drug-target interactions, and disease-related microbes. It also summarizes key databases and successful methodologies, serving as a guide for future research in this field.</p>Ahmet TOPRAKEsma ERYILMAZ DOGAN
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-019106991Statistical and experimental analysis of machining parameters and tool radius in turning of low carbon steels
https://as-proceeding.com/index.php/ijanser/article/view/2836
<p>In this study, the effects of basic machining parameters and insert radius on machinability <br>during turning of low-carbon steels were investigated using experimental and statistical methods. The <br>experiments were conducted on AISI 1015 case-hardened steel, and speed, depth, feed, and insert radius <br>were considered as input parameters. Output parameters were determined as roughness and cutting force. <br>The full factorial method was used for the experimental design, and the obtained data were evaluated <br>using ANOVA. The results show that feed and insert radius are particularly decisive on roughness. At <br>low feed and high cutting speed, a 0.8 mm insert radius provides optimum surface quality. Cutting force <br>increases in direct proportion to depth of cut and feed, while decreasing with increasing speed. It was also <br>observed that tools with a 0.4 mm insert radius exhibit lower force values. According to the ANOVA <br>results, feed was the parameter that most affected surface roughness at a 0.4 mm tip radius, while depth <br>was the most critical parameter determining force at a 0.8 mm tip radius. Consequently, this study <br>provides guidance in selecting optimal cutting parameters to achieve high efficiency and desired <br>tolerances in manufacturing processes.</p>Necmettin AydınHakan YurtkuranMustafa KuntoğluRüstem Binali
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-0191092100Türkiye’nin Hava Yolu İhracatında Sosyal ve Çevresel Sürdürülebilirlik için Yapay Zekâ Destekli Çok Amaçlı Optimizasyon Yaklaşımı
https://as-proceeding.com/index.php/ijanser/article/view/2837
<p>Bu çalışma, Türkiye’nin hava yolu ihracatında sürdürülebilirliği hem sosyal hem de çevresel <br>boyutlarıyla değerlendiren çok amaçlı bir optimizasyon yaklaşımı sunmaktadır. Doğrusal programlama <br>modeli, istihdamı artırmayı ve CO₂ emisyonlarını sınırlandırmayı amaçlamaktadır. 2000–2023 yıllarına ait <br>TÜİK ihracat verileri, ülkeler arası hava mesafeleri ve dört farklı uçak tipine /taşımacılık türüne (Belly<br>hold taşımacılığı, Boeing 747F, Boeing 777F ve Airbus A330F) ilişkin katsayılar değerlendirilmiştir. <br>Modelin çözümünde PuLP kütüphanesi kullanılmıştır. TÜİK'ten elde edilen 2000-2023 yıllarına ait veriler <br>yapay sinir ağına aktarılmış ve gelecek yılların (2024-2025) tahmini için modelden yararlanılmıştır. <br>Bulgular, istihdamın artış eğiliminde olduğunu göstermektedir. Emisyonların ise dönemsel dalgalanmalar <br>sergilediğini göstermektedir. Tahmin sonuçlarına göre, 2024 ve 2025 yılları için istihdamda yaklaşık %4,1 <br>oranında bir artış olduğu ve CO₂ emisyonlarında ise %5,0 düzeyinde bir yükseliş yaşanacağı <br>öngörülmektedir. Bu çalışma, havayolu taşımacılığında toplumsal faydalar ile çevresel etkilerin ele <br>alınmasına imkân tanımaktadır ve karar vericiler için de oldukça önemli uygulanabilir stratejik öneriler <br>sunmaktadır. Ek olarak, bu araştırma hava yolu taşımacılığında sürdürülebilir kalkınma hedeflerine <br>ulaşmak için yol gösterici bir çerçeve ortaya koymaktadır. Çalışma, çevresel maliyetlerin azaltılmasıyla <br>birlikte, toplumsal faydanın artmasına katkı sağlayacak politikaların tasarlanmasına destek olması <br>beklenmektedir. Özellikle çevresel verimlilik ile sosyal kazanımların dengelenmesi gerektiğine vurgu <br>yapılmakta ve mevcut uygulamaların yanı sıra gelecek kuşakların ihtiyaçlarını gözeten stratejik adımlar <br>için de sağlam bir temel sunulmaktadır.</p>Nesrin KolukırıkBahadır Kopçasız
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-01910101108Personal data ethics in forensic informatics: Developing innovative solution methods with TRIZ method
https://as-proceeding.com/index.php/ijanser/article/view/2838
<p>This study addresses the ethical tensions between the protection of personal data and the need <br>for evidence integrity and rapid work in forensics through the TRIZ approach. First, four basic <br>contradictions are defined: confidentiality-access, speed-procedurality, scope-proportionality, and <br>transparency-operational security. Then, feasible solutions are proposed using TRIZ's practical principles. <br>Selective acquisition and automated masking, focusing only on event-relevant data, an ethical compliance <br>matrix based on objective–data mapping. Evidence quarantine and recorded access for sensitive content. <br>Role-based authorizations that are time-bound and can be easily revoked when needed. Records that track <br>all transactions, maintaining integrity. This approach reduces unnecessary personal data exposure, <br>maintains the integrity of evidence, makes decisions visible, and facilitates auditing. Qualitative evaluations <br>show that selective acquisition and masking enable data minimization while accelerating review. It shows <br>that approval and registration flows strengthen defensibility. Proper tool support, team training, and clear <br>corporate policies are important for successful implementation. In the future, it is recommended to develop <br>evaluations with measurable metrics (completion time, mask opening rate, number of access requests), tool <br>integrations, and training programs across different types of cases. This framework offers a streamlined and <br>scalable roadmap that makes it easier to achieve operational goals while adhering to ethical principles.</p>Semih BalcıErhan AKBAL
Copyright (c) 2025 International Journal of Advanced Natural Sciences and Engineering Researches
2025-10-012025-10-01910109114