Makine Öğrenmesi Tabanlı Adi Fiğ (Vicia Sativa) Tohumlarının Fiziksel Özelliklerinin Tahmini ve Sınıflandırılması
Özet
Bu çalışmada, ekonomik açıdan önemli bir yem bitkisi olan adi fiğin (Vicia sativa L.) Cumhuriyet-99 ve Selçuk-99 çeşitlerine ait tohumların temel fiziksel özellikleri belirlenmiş ve makine öğrenmesi algoritmalarıyla tahmini ve sınıflandırılması yapılmıştır. Temmuz 2022'de Kayseri'de hasat edilen tohumlardan her çeşit için 100 adet örneklenerek dijital kumpasla uzunluk, genişlik ve kalınlık ölçümleri gerçekleştirilmiş; bu verilerden hacim, projeksiyon alanı, şekil indeksi ve uzama hesaplanmıştır. Elde edilen 1400 veri min-maks normalizasyonunun ardından %66 eğitim ve %34 test olarak ayrılarak Çok Katmanlı Algılayıcı Yapay Sinir Ağı (MLP), k-En Yakın Komşu (k-NN) ve Rastgele Orman (RF) algoritmaları ile değerlendirilmiştir. Fiziksel ölçümlerde Selçuk-99 çeşidi; uzunluk (5,74 mm), genişlik (5,06 mm), kalınlık (4,43 mm), hacim (67,58 mm³) ve projeksiyon alanı (20,02 mm²) bakımından Cumhuriyet-99'a göre anlamlı derecede yüksek değerler gösterirken, şekil indeksi ve uzama açısından çeşitler arasında istatistiksel fark saptanmamıştır. Sınıflandırma performansında %97,06 doğruluk, 0,94 kappa ve 0,986 ROC alanı değerleriyle en başarılı algoritma k-NN olmuştur. Özellik tahmininde ise tüm parametrelerde en yüksek belirleme katsayısı ($R^2$ > 0,9890) ve en düşük hata oranlarını sunan MLP algoritması öne çıkmıştır. Sonuçlar, makine öğrenmesinin fiğ tohumlarının sınıflandırılması ve niteliklerinin kestiriminde pratik, objektif ve güvenilir bir yöntem sunduğunu kanıtlamaktadır.
In this study, the primary physical properties of seeds belonging to the Cumhuriyet-99 and Selçuk-99 cultivars of common vetch (Vicia sativa L.), an economically important forage crop, were determined and predicted as well as classified using machine learning algorithms. Sampling 100 seeds per cultivar from crops harvested in Kayseri in July 2022, measurements of length, width, and thickness were conducted using a digital caliper, from which volume, projection area, shape index, and elongation were calculated. Following min-max normalization, a total of 1400 data points were partitioned into 66% training and 34% testing sets and evaluated using Multilayer Perceptron Artificial Neural Network (MLP), k-Nearest Neighbors (k-NN), and Random Forest (RF) algorithms. In physical measurements, the Selçuk-99 cultivar exhibited significantly higher values than Cumhuriyet-99 in terms of length (5.74 mm), width (5.06 mm), thickness (4.43 mm), volume (67.58 mm³), and projection area (20.02 mm²), whereas no statistically significant difference was detected between cultivars regarding shape index and elongation. In terms of classification performance, k-NN emerged as the most successful algorithm, achieving an accuracy of 97.06%, a kappa statistic of 0.94, and an ROC area of 0.986. For feature prediction, the MLP algorithm excelled across all parameters, providing the highest coefficient of determination ($R^2$ > 0.9890) and the lowest error rates. The results demonstrate that machine learning offers a practical, objective, and reliable approach for classifying vetch seeds and predicting their characteristics.
Referanslar
Abbeddou, S., Rihawi, S., Hess, H.D., Iniguez, L., Mayer, A.C., Kreuzer, M., 2011. Nutritional composition of lentil straw, vetch hay, olive leaves, and saltbush leaves and their digestibility as measured in fat-tailed sheep. Small Ruminant Research, 96: 126–135.
Açıkgöz, E., 2011. Yem Bitkileri Yetiştiriciliği. Süt Hayvancılığı, Eğitim Merkezi Yayınları Hayvancılık Serisi. Türkiye.
Akpinar, N., Akpinar, M.A., Turkoglu, S., 2001. Total lipid content and fatty acid composition of the seeds of some Vicia L. species. Food Chemistry, 74: 449–453.
Altuntas, E., Karadag, Y., 2006. Some physical and mechanical properties of sainfoin (Onobrychis sativa Lam.), grasspea (Lathyrus sativus L.) and bitter vetch (Vicia ervilia (L.) Willd.) seeds. Journal of Applied Sciences, 6: 13731379.
Berhane, T., Lane, C.,Wu, Q., Autrey, B., Anenkhonov, O., Chepinoga, V., Liu, H., 2018. Decision-tree, rule-based, and random forest classification of high-resolution multispectral imagery for wetland mapping and inventory. Remote Sensing, 10: 580.
Bradley, A.P., 1997. The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recognition, 30(7): 1145–1159.
Breiman, L., 2001. Random Forests. Machine Learning, 45(1): 5–32.
Concha-Meyer, A., Eifert, J., Wang, H., Sanglay, G., 2018. Volume estimation of strawberries, mushrooms, and tomatoes with a machine vision system. International Journal of Food Properties, 21(1): 1867-1874.
Costa, C., Antonucci, F., Pallottino, F., Aguzzi, J., Sun, D.W., Menesatti, P., 2011. Shape analysis of agricultural products: a review of recent research advances and potential application to computer vision. Food and Bioprocess Technology, 4: 673–692.
Çaçan, E., Kökten, K., Kaplan, M., Yılmaz, Ş.H., 2018. Bazı Adi Fiğ hat ve çeşitlerinin (Vicia sativa L.) ot verimi ve ot kalitesi açısından değerlendirilmesi. Harran Tarım ve Gıda Bilimleri Dergisi, 22(1): 47-61
Çetin, N., Sağlam, C., 2022. Rapid detection of total phenolics, antioxidant activity and ascorbic acid of dried apples by chemometric algorithms. Food Bioscience, 47: 101670.
Çetin, N., Karaman, K., Beyzi, E., Sağlam, C., Demirel, B., 2021. Comparative evaluation of some quality characteristics of sunflower oilseeds (Helianthus annuus l.) through machine learning classifiers. Food Analytical Methods, 14: 1666–1681
Çetin, N., 2022. Machine learning for varietal binary classification of soybean (Glycine max (L.) Merrill) seeds based on shape and size attributes. Food Analytical Methods, 15: 2260–2273.
Dumanoğlu, Z., Çaçan, E., Kökten, K., 2022. Yayın fiğ (Vicia sativa L.) tohumlarına ait bazı morfolojik ve fizyolojik özelliklerinin belirlenmesi üzerine bir araştırma. MAS Journal of Applied Sciences, 7(1): 41-47.
Ford, R., Maddeppungeng, A.M., Taylor, P.W.J., 2008. Vetch. In: C. Kole, T. C. Hall (eds), Compendium of transgenic crop plants: Transgenic legume seeds and forages, Vol 9. Blackwell Publishing, Oxford, pp. 163–170.
Hall, M., Frank, E., Holmes, G., Pfahringer, B., Reutemann, P., Witten, I.H., 2009. The WEKA data mining software: an update. SIGKDD Explorations.
Işık, E., İzli, N., 2016. Effects of moisture content on some physical properties of the yellow lentil. Journal of Agricultural Sciences, 22(2): 307-316.
Karray, F., Karray, F.O., De Silva, C.W., 2004. “Soft Computing and Intelligent Systems Design: Theory, Tools, and Applications”. Pearson Education.
Khazaei, J., Jafari, S., Noorolah, S., 2008. Lognormal vs. Normal and Weibull distributions for modeling the mass and size distributions of sunflower seeds and kernels. World Conference on Agricultural Information and IT.
Kibar, H., Öztürk, T., Temizel, K.E., 2014. Effective engineering properties in the design of storage structures of postharvest dry bean grain. Acta Scientiarum Agronomy, 36: 147-158.
Larbi, A., El-Moneim, A.M., Nakkoul, H., Jammal, B., Hassan, S., 2011. Intra-species variations in yield and quality determinants in Vicia species: 3. common vetch (Vicia sativa ssp. sativa L.). Animal Feed Science and Technology, 164: 241–251.
Mao, Z., Fu, X., Nan, Z.B.H., Wan, C.G., 2015. Fatty acid, amino acid, and mineral composition of four common vetch seeds on Qinghai-Tibetan plateau. Food Chemistry, 171: 13–18.
Marini, F., Magri, A.L., Balestrieri, F., Fabretti, F., Marini, D., 2004. Supervised pattern recognition applied to the discrimination of the floral origin of six types of Italian honey samples. Analytica Chimica Acta, 515(1):117–125.
Maxwell, A.E., Warner, T.A., Fang, F., 2018. Implementation of machine learning classification in remote sensing: an applied review. International Journal Remote Sensing, 39: 2784–2817.
Mikic, A., Peric, V., Dordevi Srebric, M., Mihailovic, V., 2009. Anti-nutritional factors in some grain legumes. Biotechnology in Animal Husbandry, 25: 1181–1188.
Mollazade, K., Omid, M., Arefi, A., 2012. Comparing data mining classifiers for grading raisins based on visual features. Computers and Electronics in Agriculture, 84:124–131.
Nettleton, D.F., Orriols-Puig, A., Fornells, A., 2010. A study of the effect of different types of noise on the precision of supervised learning techniques. Artificial Intelligence Review, 33: 275–306.
Omid, M., Khojastehnazhand, M., Tabatabaeefar, A., 2010. Estimating volume and mass of citrus fruits by image processing technique. Journal Food Engineering, 100(2): 315–321.
Pietersma, D., Lacroix, R., Lefebvre, D., Wade, K.M., 2003. Performance analysis for machine-learning experiments using small data sets. Computers Electronics in Agriculture, 38(1): 1–17.
Pinkerton, B., Pinkerton, F., 2000. Managing forages for meat goats. In: Meat goat production handbook. Extention services. Collage of Agriculture, Forestry and Life Science, Clemson University, USA.
Ponce, J.M., Aquino, A., Millán, B., Andújar, J.M., 2018. Olive-fruit mass and size estimation using image analysis and feature modeling. Sensors, 18(9): 2930.
Rodriguez-Galiano, V.F., Ghimire, J., Rogan, B., Chica-Olmo, M., Rigol-Sanchez, J.P., 2012. An assessment of the effectiveness of a Random Forest classifier for land-cover classification. ISPRS J Photogramm Remote Sensing, 67: 93–104.
Romero, J.R., Roncallo, P.F., Akkiraju, P.C., Ponzoni, I., Echenique, V.C., Carballido, J.A., 2013. Using classification algorithms for predicting durum wheat yield in the province of Buenos Aires. Computers Electronics in Agriculture, 96: 173–179.
Sağlam, C., Çetin, N., 2022. Machine learning algorithms to estimate drying characteristics of apples slices dried with different methods. Journal of Food Processing and Preservation, e16496.
Saha, K.K., Uddin, M.Z., Rahman, M.M., Moniruzzaman, M., Ali, M.A., Oliver, M.M.H., 2021. Estimation of cardamom capsule size and surface area using digital image processing technique. Journal of the Bangladesh Agricultural University, 19(3): 398-405.
Sahay, K.M., Singh, K.K., 2004. A textbook on unit operation of agricultural processing.
Singh, S.K., Vidyarthi, S.K., Tiwari, R., 2020. Machine learnt image processing to predict weight and size of rice kernels. Journal of Food Engineering, 274: 109828.
Stegmayer, G., Milone, D.H., Garran, S., Burdyn, L., 2013. Automatic recognition of quarantine citrus diseases. Expert Systems Applications, 40(9): 3512–3517.
Tabatabaeefar, A., Rajabipour, A., 2005. Modeling the mass of apples by geometrical attributes. Science Horticulture, 105(3): 373–382
Taser, O.F., Altuntas, E., Ozgoz, E., 2005. Physical properties of Hungarian and common vetch seeds. Journal of Applied Sciences, 5(2): 323-326.
Varol, İ.S., Çetin, N., Kırnak, H., 2022. Comparative evaluation of quality properties of chickpea seed (Cicer arietinum L.) grown under different irrigation conditions by using machine learning. Journal of Agricultural Sciences, 1: 12.
Wang, X., Zhang, H., Song, R., He, X., Mao, P., Jia, S., 2021. Non-destructive identification of naturally aged alfalfa seeds via multispectral imaging analysis. Sensors, 21(17): 5804.
Witten, I.H., Frank, E., 2005. Data mining: practical machine learning tools and techniques. Morgan Kaufmann Press, San Francisco.
Xu, G., Shen, C., Liu, M., Zhang, F., Shen, W., 2017. A user behavior prediction model based on parallel neural network and k-nearest neighbor algorith ms. Cluster Computing, 20(2): 1703-1715.
Xu, P.,Yang, R., Zeng, T., Zhang, J., Zhang, Y., Tan, Q., 2021. Varietal classification of maize seeds using computer vision and machine learning techniques. Journal of Food Process Engineering, 44(11): e13846.
Yalçın, İ., Özarslan, C., 2004. Physical properties of vetch seed. Biosystems Engineering, 88(4): 507-512.
Yang, L., Zhang, Z., Hu, X., 2020. Cultivar discrimination of single alfalfa (Medicago sativa L.) seed via multispectral imaging combined with multivariate analysis. Sensors, 20(22): 6575.
Zhang, H., Song, T., Wang, K., Wang, G., Hu, H., Zeng, F., 2012. Prediction of crude protein content in rice grain with canopy spectral reflectance. Plant, Soil and Environment, 58: 514–520.
