Journals / Dokuz Eylül Üniversitesi Mühendislik Fakültesi Fen ve Mühendislik Dergisi / 2026 / Cilt: 28 Sayı: 83

Benchmarking Machine Learning and Transfer Learning Approaches for Fruit Classification Using Explainable Artificial Intelligence

Açıklanabilir Yapay Zeka Kullanılarak Meyve Sınıflandırması için Makine Öğrenmesi ve Transfer Öğrenme Yaklaşımlarının Kıyaslanması

Gözde Alp(AYDIN ADNAN MENDERES UNIVERSITY)0000-0002-6479-3500
Fatih Soygazi(AYDIN ADNAN MENDERES UNIVERSITY)0000-0001-8426-2283
Pages
302–310
Publication Date
2026-05-31
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

The results underline the significance of both model capability and explainability when it comes to model selection, and show that fine-tuned deep networks combined with explainability tools is a strong and useful framework for image based classification.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:In this study, we investigate traditional machine learning methods with respect to state-of-the-art deep learning approaches for fruit image classification.
📊
Key Findings & Data:Conventional classifiers such as Random Forest, and eXreme Gradient Boosting were evaluated alongside neural networks and state-of-the-art transfer learning models, including Residual Network (ResNet), Visual Geometry Group 16 (VGG16), and EfficientNet-B0.
💡
Conclusion & Impact:The results underline the significance of both model capability and explainability when it comes to model selection, and show that fine-tuned deep networks combined with explainability tools is a strong and useful framework for image based classification.
🔬Design:Modeling & Simulation
📍Scope:Comprehensive Sample
⚙️Variables:Derin Öğrenme · Açıklanabilir Yapay Zeka · GradCAM

Abstract

In this study, we investigate traditional machine learning methods with respect to state-of-the-art deep learning approaches for fruit image classification. Conventional classifiers such as Random Forest, and eXreme Gradient Boosting were evaluated alongside neural networks and state-of-the-art transfer learning models, including Residual Network (ResNet), Visual Geometry Group 16 (VGG16), and EfficientNet-B0. The fine-tuned ResNet50 model obtained the maximum classification accuracy as 97.57%, substantially surpassing other traditional solutions as well as shallow networks. One important contribution of this work is in the focus on model explainability. To understand the reasoning from the deep models, we used Gradient-weighted Class Activation Mapping (Grad-CAM) and Grad-CAM++ techniques to visualize what kind of decision is helping the model making predictions as well as by considering what regions in input images are being looked up for the concerned class. This explainability integration provides the transparency and explainability of classification system, being crucial for its real world applicability. The results underline the significance of both model capability and explainability when it comes to model selection, and show that fine-tuned deep networks combined with explainability tools is a strong and useful framework for image based classification.

Özet

Bu çalışmada, meyve görüntüsü sınıflandırması için geleneksel makine öğrenmesi algoritmaları ve modern derin öğrenme tekniklerinin karşılaştırmalı bir analizini sunuyoruz. Random Forest ve eXreme Gradient Boosting gibi geleneksel sınıflandırıcılar, sinir ağları ve Residual Network (ResNet), Visual Geometry Group 16 (VGG16) ve EfficientNet-B0 gibi en güncel transfer öğrenme modelleri ile birlikte değerlendirildi. Bunlar arasında, ince ayarlı ResNet50 mimarisi, %97.57'lik en yüksek sınıflandırma doğruluğunu elde ederek hem geleneksel yaklaşımları hem de sığ ağları açıkça geride bırakmıştır. Bu çalışmanın temel katkılarından biri, model açıklanabilirliğine verilen önemdir. Derin modellerin karar verme sürecini görselleştirmek için Gradyan Ağırlıklı Sınıf Aktivasyon Eşlemesi (Grad-CAM) ve Grad-CAM++ tekniklerini kullandık ve giriş görüntülerindeki ilgi bölgelerine dair içgörüler sunduk. Açıklanabilirliğin bu şekilde bütünleştirilmesi, özellikle gerçek dünyada çalıştırma veya kullanım senaryolarında kritik öneme sahip olan sınıflandırma sisteminin şeffaflığını ve güvenilirliğini artırır. Bulgular, model seçiminde hem doğruluğun hem de yorumlanabilirliğin önemini vurgulamaktadır ve ince ayarlı derin ağların açıklanabilirlik araçlarıyla birleştirilmesinin, görüntü tabanlı sınıflandırma görevleri için sağlam ve bilgilendirici bir çerçeve sağladığını öne sürmektedir.

🔖

Cite This Article

Alp, G., Soygazi, F. (2026). Benchmarking Machine Learning and Transfer Learning Approaches for Fruit Classification Using Explainable Artificial Intelligence. Dokuz Eylül Üniversitesi Mühendislik Fakültesi Fen ve Mühendislik Dergisi, Cilt: 28 Sayı: 83, 302–310. https://doi.org/10.21205/deufmd.2026288316

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (23)

  1. Song X, Zhang X, Dong G, Ding H, Cui X, Han Y, et al. AI in food industry automation: applications and challenges. Frontiers in Sustainable Food Systems 2025;9. doi:10.3389/fsufs.2025.1575430.
  2. Thapa A, Nishad S, Biswas D, Roy S. A comprehensive review on artificial intelligence assisted technologies in food industry. Food Bioscience 2023;56. doi:10.1016/j.fbio.2023.103231.
  3. Zhu L, Spachos P, Pensini E, Plataniotis KN. Deep learning and machine vision for food processing: A survey. Current Research in Food Science 2021;4:233-249. doi:10.1016/j.crfs.2021.03.009.
  4. Muresan H, Oltean M. Fruit recognition from images using deep learning. Acta Universitatis Sapientiae, Informatica 2018;10(1):26-42. doi:10.2478/ausi-2018-0002.
  5. Ukwuoma CC, Zhiguang Q, Bin Heyat MB, Ali L, Almaspoor Z, Monday HN. Recent advancements in fruit detection and classification using deep learning techniques. Mathematical Problems in Engineering 2022;2022(1):9210947. doi:10.1155/2022/9210947.
  6. Salim NO, Mohammed AK. Comparative Analysis of Classical Machine Learning and Deep Learning Methods for Fruit Image Recognition and Classification. Traitement du Signal 2024;41(3). doi:10.18280/ts.410322.
  7. Zhang D. Fruit 360 classification based on the convolutional neural network. Applied and Computational Engineering 2023;15:219-222.
  8. Joseph JL, Kumar VA, Mathew SP. Fruit classification using deep learning. In: Innovations in Electrical and Electronic Engineering: Proceedings of ICEEE 2021. Springer Singapore; 2021, p. 807-817.
  9. Sinha BB, Dhanalakshmi R. A multi-fused convolutional neural network model for fruit image classification. International Journal of Cognitive Computing in Engineering 2024;5:416-424. doi:10.1016/j.ijcce.2024.09.003.
  10. Bongulwar DM, Talbar SN. Robust Convolutional Neural Network Model For Recognition of Fruits. Indian Journal of Science and Technology 2021;14(45):3318-3334. doi:10.17485/IJST/v14i45.1493.
  11. Siddiqi R. Comparative performance of various deep learning based models in fruit image classification. In: Proceedings of the 11th International Conference on Advances in Information Technology; 2020, p. 1-9. doi:10.1145/3406601.3406619.
  12. Siddiqi R. Fruit-classification model resilience under adversarial attack. SN Applied Sciences 2022;4(1):31. doi:10.1007/s42452-021-04917-6.
  13. Zárate V, Hernández DC. Simplified Deep Learning for Accessible Fruit Quality Assessment in Small Agricultural Operations. Applied Sciences 2024;14(18):8243. doi:10.3390/app14188243.
  14. Rathnayake N, Rathnayake U, Dang TL, Hoshino Y. An efficient automatic fruits-360 image identification and recognition using a novel modified cascaded-ANFIS algorithm. Sensors 2022;22(12):4401. doi:10.3390/s22124401.
  15. Mimma NEA, Ahmed S, Rahman T, Khan R. Fruits classification and detection application using deep learning. Scientific Programming 2022;2022(1):4194874. doi:10.1155/2022/4194874.
  16. Altaheri H, Alsulaiman M, Muhammad G, Amin SU, Bencherif M, Mekhtiche M. Date fruit dataset for intelligent harvesting. Data in Brief 2019;26:104514. doi:10.1016/j.dib.2019.104514.
  17. Abayomi-Alli OO, Damaševičius R, Misra S, Abayomi-Alli A. FruitQ: a new dataset of multiple fruit images for freshness evaluation. Multimedia Tools and Applications 2024;83(4):11433-11460. doi:10.1007/s11042-023-16058-6.
  18. Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Scikit-learn: Machine learning in Python. The Journal of Machine Learning Research 2011;12:2825-2830.
  19. Wang SH, Chen Y. Fruit category classification via an eight-layer convolutional neural network with parametric rectified linear unit and dropout technique. Multimedia Tools and Applications 2020;79(21):15117-15133. doi:10.1007/s11042-018-6661-6.
  20. Marmolejo-Saucedo JA, Kose U. Numerical grad-cam based explainable convolutional neural network for brain tumor diagnosis. Mobile Networks and Applications 2024;29(1):109-118. doi:10.1007/s11036-022-02021-6.
  21. Hu B, Tunison P, RichardWebster B, Hoogs A. Xaitk-saliency: An open source explainable ai toolkit for saliency. In: Proceedings AAAI Conference on Artificial Intelligence; 2023;37(13):15760-15766. doi:10.1609/aaai.v37i13.26871.
  22. Benedichuk M, Bashkova P, Tuchinov B. Computer Vision and Explainable Approaches for Chest Tuberculosis Screenings. In: 2024 IEEE 3rd International Conference on Problems of Informatics, Electronics and Radio Engineering (PIERE). IEEE; 2024, p. 820-823. doi:10.1109/piere62470.2024.10804967.
  23. Nahshan Y, Chmiel B, Baskin C, Zheltonozhskii E, Banner R, Bronstein AM, Mendelson A. Loss aware post-training quantization. Machine Learning 2021;110(11):3245-3262. doi:10.1007/s10994-021-06053-z.