Multi-Class Document Classification Based on Deep Neural Network and Word2Vec

Metin Zontul

Makale | 2022 | JOURNAL OF AERONAUTICS AND SPACE TECHNOLOGIES15 ( 1 ) , pp.59 - 65

With the increase in unstructured data, the importance of classification of text-based documents has increased. In particular, the classification of news texts and digital documentation provides easy access to the information sought. In this study, a large amount of news textual data was used. After the data set was preprocessed, Bag of Words (BoW), TF-IDF, Word2Vec and Doc2Vec word embedding methods were applied. In the classification phase, Random Forest (RF), Multilayer Perceptron (MLP), Support Vector Machine (SVM) and Deep Neural Network (DNN) algorithms were applied. As a result of the experimental studies, using the Word2Vec . . .method together with the DNN algorithm performed the best result. Yapısal olmayan verilerin artmasıyla birlikte metin tabanlı belgelerin sınıflandırılmasının önemi artmıştır. Özellikle haber metinlerinin sınıflandırılması ve dijital dokümantasyon, aranan bilgilere kolay erişim sağlar. Bu çalışmada, büyük miktarda metinsel haber verisi kullanılmıştır. Veri seti ön işlemeye tabi tutulduktan sonra, Bag of Words (BoW), TF-IDF, Word2Vec ve Doc2Vec kelime temsil yöntemleri uygulanmıştır. Sınıflandırma aşamasında Random Forest (RF), Multilayer Perceptron (MLP), Support Vector Machine (SVM) ve Deep Neural Network (DNN) algoritmaları uygulanmıştır. Deneysel çalışmalar sonucunda DNN algoritması ile birlikte Word2Vec yönteminin kullanılması en iyi sonucu vermiştir Daha fazlası Daha az

Web Service-Based Two-Dimensional Vehicle Pallet Loading with Routing for a Real-World Problem

Metin Zontul

Makale | 2022 | Conference Proceedings

Since increasing oil prices and vehicle costs increase transportation costs in order delivery systems, the optimal vehicle loading and routing is very crucial for the companies in competitive conditions. Although there are many studies related to optimal vehicle loading and routing by using linear programming and heuristic algorithms, there is not enough practical web service-based application in the literature. In this study, we propose a hybrid model to solve the problem of two-dimensional vehicle pallet loading with routing for a real-world data by combining Knapsack Problem solver algorithms such as MaxRects, Skyline and Guillot . . .ine with Dijkstra's algorithm for loading and routing respectively as a web service-based application. © 2022 IEEE Daha fazlası Daha az

A Smart and Mechanized Agricultural Application: From Cultivation to Harvest

Metin Zontul

Makale | 2022 | APPLIED SCIENCES-BASEL ( 12 ) , pp.153 - 156

Food needs are increasing day by day, and traditional agricultural methods are not responding efficiently. Moreover, considering other important global challenges such as energy sufficiency and migration crises, the need for sustainable agriculture has become essential. For this, an integrated smart and mechanism-application-based model is proposed in this study. This model consists of three stages. In the first phase (cultivation), the proposed model tried to plant crops in the most optimized way by using an automized algorithmic approach (Sand Cat Swarm Optimization algorithm). In the second stage (control and monitoring), the gro . . .wing processes of the planted crops was tracked and monitored using Internet of Things (IoT) devices. In the third phase (harvesting), a new method (Reverse Ant Colony Optimization), inspired by the ACO algorithm, was proposed for harvesting by autonomous robots. In the proposed model, the most optimal path was analyzed. This model includes maximum profit, maximum quality, efficient use of resources such as human labor and water, the accurate location for planting each crop, the optimal path for autonomous robots, finding the best time to harvest, and consuming the least power. According to the results, the proposed model performs well compared to many well-known methods in the literature Daha fazlası Daha az

A local-holistic graph-based descriptor for facial recognition

Metin Zontul

Makale | 2022 | MULTIMEDIA TOOLS AND APPLICATIONS , pp.153 - 156

Face recognition remains critical and up-to-date due to its undeniable contribution to security. Many descriptors, the most vital figures used for face discrimination, have been proposed and continue to be done. This article presents a novel and highly discriminative identifier that can maintain high recognition performance, even under high noise, varying illumination, and expression exposure. By evolving the image into a graph, the feature set is extracted from the resulting graph rather than making inferences directly on the image pixels as done conventionally. The adjacency matrix is created at the outset by considering the pixel . . .s’ adjacencies and their intensity values. Subsequently, the weighteddirected graph having vertices and edges denoting the pixels and adjacencies between them is formed. Moreover, the weights of the edges state the intensity differences between the adjacent pixels. Ultimately, information extraction is performed, which indicates the importance of each vertex in the graphic, expresses the importance of the pixels in the entire image, and forms the feature set of the face image. As evidenced by the extensive simulations performed, the proposed graphic-based identifier shows remarkable and competitive performance regarding recognition accuracy, even under extreme conditions such as high noise, variable expression, and illumination compared with the state-of-the-art face recognition methods Daha fazlası Daha az

Automated evaluation of Cr-III coated parts using Mask RCNN and ML methodse

Metin Zontul | Esra Kavalci Yilmaz | Oguz Kaynar | Ramazan Katirci


In this study, chrome coatings were carried out using a Cr-III electroplating bath. The coated parts were classified depending on their appearance. A new approach was developed to classify the coated parts automatically using artificial intelligence methods. Mask RCNN and machine learning (ML) methods such as Multilayer Perceptron (MLP), Support Vector Classifier (SVC), Gaussian Process (GP), K-nearest Neighbors (KNN), XGBoost, and Random Forest Classifier (RFC) were used together. Mask RCNN was used to clean the coated parts from the redundant data. The extracted data were flattened and converted to the row vectors for use as input . . . in ML methods. ML algorithms were used to classify the coated parts as Pass and Fail. The classification accuracy was checked with the leave one out (loo) cross-validation method. RFC method gave the highest accuracy, 0.83, and F1 score, 0.88. The accuracy of Mask RCNN was checked using a dataset of separated validation images. It was observed that extracting the unnecessary data from the images increased the accuracy exceedingly. Moreover, the method exhibits a high potential to keep the parameters of the electroplating process under control Daha fazlası Daha az

6698 sayılı Kişisel Verilerin Korunması Kanunu kapsamında yükümlülüklerimiz ve çerez politikamız hakkında bilgi sahibi olmak için alttaki bağlantıyı kullanabilirsiniz.

Bu site altında yer alan tüm kaynaklar Creative Commons Alıntı-GayriTicari-Türetilemez 4.0 Uluslararası Lisansı ile lisanslanmıştır.