Authors:Seçmen Şahin; Güneş Harman Abstract: Quantum Convolutional Neural Networks (QCNNs) aim to enhance the capabilities of convolutional neural networks by leveraging the strengths of quantum computing. They operate by locally transforming input data using quantum circuits. In this study, two models have been built on a quantum-encoded COVID-19 dataset. Model-1 classifies between ‘Normal Person’ and ‘Covid19/Viral Pneumonia’, while Model-2 classifies between ‘Covid-19’ and ‘Viral Pneumonia’. Three different classifications have been made based on the number of qubits (feature count) for these models. For Quantum Classifier 1, approximately 70% accuracy was achieved by extracting 11 features from the 256-feature input data obtained through basic data analysis. In Quantum Classifier 2, using the TruncatedSVD method, each image’s 256 features were reduced to 4, resulting in 72% accuracy. Finally, Quantum Classifier 3 achieved an unexpected 76% accuracy using only 2 features. These models provide significant insights into diagnosing diseases from lung X- Ray images and demonstrate how quantum computers can be effectively utilized in the healthcare domain. Additionally, the impact of different parameters in the “default qubit” device of Pennylane on model performance has been investigated. The study highlights how Quantum Classifier 3 achieves high accuracy by significantly reducing the data dimension, indicating the potential of QCNNs to provide high performance with less resource usage in the future. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:Onur Karaçay; Süleyman Kılıç Abstract: Tractor tires, thanks to their large surface areas, reduce the pressure applied to the soil, thus minimizing soil compaction and thereby increasing efficiency. Technological advancements, including increased engine power, have led to larger and heavier tractors, further increasing the risk of soil compaction. With these developments, new tires with low inflation pressures have been developed to reduce the surface pressure on the soil. These new tires are referred to as increased flexion (IF) and very high flexion tires. In this study, a comparison was made between a normal structure 600/70 R30 tire and an IF structure IF 600/70 R30 tire under a fixed load with three different rim structures, examining changes in pressure as well as changes in the footprint and deflection. The results show that at the minimum inflation pressure of 12psi, both tires exhibited the maximum ground contact area, whereas the minimum contact area was achieved at the maximum inflation pressure of 35psi with a DW21 rim for both tires. The maximum deflection value was recorded at 12psi with a W18 rim. It has been determined that the choice of rim and inflation pressure affects the contact area of the tire, and to reduce the surface pressure on the soil, it is necessary to reduce the tire inflation pressure to increase the contact area. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:Mustafa Alptekin Engin Abstract: Parkinson’s disease (PD), a neurological disorder, negatively affects the lives of patients and their caregivers. PD, which is very difficult to diagnose early by examining the clinical characteristics of the person, can be diagnosed using voice recordings. However, the inconsistent performance results of the models obtained from the evaluation of voice recordings through machine learning techniques limit the usability of these models to aid in diagnosing physicians. This study used a database of 195 voice data obtained from 31 individuals, 23 of whom have PD. The classification of the voices as healthy or patient was based on the 22 features in the database. The split ratios 90/10, 80/20, 70/30, 50/50 and 30/70 were used to select the training and test phase data, respectively. In addition, each split ratio was evaluated using 10-fold cross-validation, 5-fold cross-validation, holdout validation and resubstitution validation methods in the training phase, which is the initial process that will directly affect the other classification procedures. In addition, the classification process was performed using quadratic discriminant analysis, support vector machine, ensemble bagged tree, k-nearest neighbours and neural network classifiers. All procedures were repeated 10 times to ensure consistency of results and randomisation of split ratios. As a result, the k-nearest neighbours classifier with 80/20 splitting ratio and 10-fold cross-validation was determined to be the most successful among the compared methods with 95.64±3.21% accuracy. Therefore, it can be seen that much more successful results can be obtained by analysing only the effects of the existing parameters of the classifiers. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:Fatma Zehra Solak Abstract: Knee Osteoarthritis (KOA) is the most common type of arthritis and its severity is assessed with the Kellgren-Lawrence (KL) grading system based on evidence from both knee bones. Recent advancements point to an era where computer-assisted methods enhance KOA diagnostic efficiency. This study implemented binary and multiple classification processes based on X-ray images and deep learning algorithms for computer-aided KOA severity diagnosis. Pre-processing involved extracting the region of interest and contrast enhancement with CLAHE on the X-ray images from the included dataset. Using this dataset, 2, 3, 4, and 5 class classification processes were conducted with ResNet-50, Xception, VGG16, EfficientNetb0, and DenseNet201 transfer learning models. Each model was assessed with “rmsprop,” “sgdm,” and “adam” optimization algorithms. Study findings reveal that, the DenseNet201-rmsprop model achieved 87.7% accuracy, 87.2% F1-Score, and a 0.75 Cohen’s kappa value for 2-class classification. For 3-class classification, it achieved 85.6% accuracy, 82.4% F1-Score, and a 0.71 Cohen’s kappa value. For 4-class classification, the DenseNet201-rmsprop model provided 81.5% accuracy, 77.1% F1-Score, and a Cohen’s kappa value of 0.67. In the 5-class classification, the highest success was with the Xception-rmsprop model, with 67.8% accuracy, 68.8% F1-Score, and a 0.55 Cohen’s kappa value. The evaluation with varying class numbers and different transfer learning models highlights the proposed approach’s effectiveness. Results of the study underscore the study’s uniqueness and success in demonstrating how varying the number of classes, employing different transfer learning models and optimizers can provide clearer insights into KOA severity evaluation. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:Süleyman Güneş; Rıfat Hacıoğlu Abstract: Considering the sustainability problems and environmental impacts of fossil fuels, renewable and sustainable energy is becoming widespread in the world as well as in Turkey. Considering the increasing energy demand, the ability of educational institutions such as universities to produce their own energy is an important research topic in terms of the country’s economy and energy efficiency. In this study, it is aimed to create a prediction for the analysis and design of solar power plants (SPP) that may be established in order to meet the energy needs of institutions. Campus areas of universities have been selected to achieve this purpose, and installation scenarios are carried out in five different university campuses with comparisons of insolation, location and temperature change. Within the framework of current regulations, design approaches are suggested by analyzing not only radiation and sunshine duration but also panel temperature and panel angle results in energy conversion. Three different solar power plant installations are being evaluated on university campuses, including land, sloping roof and flat roof systems. In addition, energy production efficiency is also examined by examining fixed angle, monthly variable angle on a single axis and bi-annual angle changing situations. PVGIS online database and calculation system is used for analysis of solar power plant systems installed in different provinces. PVGIS is an online European Union project calculation system that provides information on solar radiation, sunshine duration, ambient conditions and photovoltaic (PV) system performance based on location information. Using PVGIS, meteorological data can be accessed for any coordinate around the world and how much energy can be produced from different PV systems can be calculated. In this design, which is compared with the real data collected, the error levels that may be encountered are revealed and the cost calculation is evaluated. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:Can Murat Dikmen; Kübra Karataş Selam Abstract: In this article, using Jacobsthal and Jacobsthal-Lucas sequences, we define generalized Jacobsthal-Like sequences and investigate their algebraic properties like Binet’s formula, generating functions, Simson formula and summation formula. We also prove some other summation formulas like sum of even and odd indices and alternating sum of generalized Jacobsthal-Like sequences. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:Berna Aksoy; İsmail Hakkı Özölçer, Emrah Doğan Abstract: During the water cycle, substances that are contaminated in water cause physical, chemical or biological alterations of the water’s natural features, therefore environmental balance deteriorate over time. Observations and measurements on a river give the necessary information about how to benefit from the river. For this reason, it is important to investigate the water quality in rivers and water reservoirs which are close to settlement areas. In this study, surface water quality measurements were carried out at five observation stations along the main line of the Filyos River, which forms the largest sub-basin in the Western Black Sea Basin, at intervals of thirty days in 2015 year. The turbidity parameter was estimated by artificial neural networks (ANNs) based on water quality parameters such as chromium (Cr+3), chemical oxygen demand (COD), iron (Fe+3), aluminium (Al+3), suspended solids, manganese (Mn+2), zinc (Zn+2), lead (Pb+2) and calcium (Ca+2). The study was conducted with creating two scenarios. In the first scenario, the determined parameters were analyzed by ANN for each station one by one. The obtained data showed that Cr (coefficient of determination [R2] =0.9999) parameter gave the best performance in the estimation of turbidity parameter in the study area. In the second scenario, eight models were created by adding the other best performing parameters one by one to the best performing Cr parameter. The third model formed by Cr, COD, Fe and Al parameters gave the closest result with R2=0.9992. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:Ali Sarı; Kamal Ismayılzada, Sinan Akıska, Fuat Erol Abstract: In this study, the redox-sensitive trace elements in the Miocene-aged bituminous claystones rich in shallow organic matter (%TOC: 14.52-44.44; avg: 31.24) in the Ilgın basin (Konya) were investigated. The elements studied include Vanadium (V), Uranium (U), and Molybdenum (Mo), as well as Zinc (Zn), Nickel (Ni), Copper (Cu), and Cobalt (Co), aiming to analyze their geochemical behaviors. Additionally, the influence of the basin’s redox conditions on organic matter accumulation was examined. For this purpose, 14 samples were systematically collected from the bituminous claystones, starting from the lignite level at the base up to the top. In the investigation of the behavior of redox-sensitive elements in the Miocene-aged bituminous claystones in the Ilgın area, the relationships between major and trace elements and their total organic carbon (%TOC) content were examined. Major and trace element analyses were conducted on samples using an ICP-OES device, and %TOC analyses were performed using a Rock Eval VI device. In the examined bituminous rocks, there are moderate correlations between %TOC and %Mo (r=0.529); weak and very weak correlations are observed for Cu (r=-0.230), Ni (r=-0.030), Zn (r=0.216), U (r=0.083), V (r=0.124), and Co (r=0.076). This indicates that the enrichment of these elements in sediments and sapropels from water masses does not involve organometallic ligands in humic acids except for Mo. Very weak correlation relationships of Fe with U (r=0.204), Ni (r=0.029), and Zn (r=-0.142) indicate that the enrichment of U and Zn in sapropel is not influenced by pyrite. However, the moderate correlation of Fe with Co (r=0.535) and strong correlation with Mo (r=0.722) suggest that the enrichment of Co and Mo in sapropel is influenced by pyrite. Very weak correlations between Mn and Cu (r=-0.562), Zn (r=-0.163), Ni (r=-0.318), V (r=-0.243), U (r=-0.142), and Mo (r=-0.600) indicate that manganese oxyhydroxides play no role in the diffusion and enrichment of these elements from water masses to sapropel. Sulfur shows very weak correlations with Ni (r=-0.121), V (r=-0.177), and Zn (r=-0.354); weak correlations with Cu (r=0.290) and U (r=0.302); moderate correlation with Co (r=0.476); and strong correlation with Mo (r=0.729). This suggests that Co precipitates as CoS and Mo as MoS2, while Cu, Zn, Ni, U, and V do not enrich as sulfides. A very strong correlation (r=0.929) between Fe and S indicates that Fe precipitates as pyrite (FeS2), suggesting an anoxic redox condition. The redox conditions in the basin are oxic/suboxic and anoxic based on Th/U ratios in all samples, except for samples RE-1 and RE-7; based on U/Th ratios, except for samples RE-1 and RE-7; and based on V/V+Ni ratios, except for samples RE-1, 2, 3, 6, and 12. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:İlker Akın; Egemen Foto Abstract: Today, metallurgy, galvanizing, leather tanning, etc. the widespread use of chromium in industries causes the release of aqueous chromium to the environment. In aqueous solutions, chromium is usually present as Cr(VI) or Cr(III). These two forms of chromium have different chemical, biological and environmental effects. Cr(VI) exists as anionic species such as HCrO4-, Cr2O72- and CrO42-, which are highly mobile in soil and water systems. In this study, the removal of Cr(VI) from aqueous solution by batch adsorption method using gum arabic modified magnetic nanoparticles (GA-MNP) was investigated. AZ-MNP was obtained by modifying the carboxylic groups of gum arabic through the interaction between the hydroxyl groups on the surface of Fe3O4. Surface modification with gum arabic led to the formation of secondary particles with diameters in the range of 9-13 nm. The adsorption rate was fast enough to reach equilibrium within 45 minutes due to the lack of internal diffusion resistance. The adsorption capacities for MNP and AZ-MNP increased with decreasing solution pH. Maximum adsorption capacity, Langmuir adsorption constant and enthalpy changes in GA-MNP were determined as 0.194 mg/g, 11.06 mg/L and 9.4 kJ/mol, respectively. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:Melike İşgören; İsmail Toroz Abstract: Rubber is a versatile material with many applications in various industrial fields. However, its extensive use also leads to the generation of significant amounts of waste. Therefore, waste reduction and the promotion of waste recovery and reuse are crucial in waste management. Therefore, accurate recognition and identification of waste is essential for proper management. This study aims to determine the physical, chemical, and technical properties of Ethylene Propylene Diene Copolymer (EPDM), a type of rubber, to identify the reuse areas of production residues from the automotive industry. The analysis results will be evaluated. The evaluation of alternatives for field use and the export of granule material produced from scrap material has been conducted under current legislation. The materials used for sealing windows and doors in the automotive industry, specifically EPDM rubber wicks, are recyclable scrap. The excess products produced during their production should also be evaluated as recyclable waste within the GTIP codes ‘400270000000: Ethylene-Propylene-Unconjugated Diene Rubber (EPDM).’ It was concluded that along with other filling materials, it is suitable for use in areas such as children’s playgrounds and sports fields for humanitarian purposes. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:Duygu Geçkin; Güleser Kalaycı Demir Abstract: A wide range of biological processes, including signal transmission, immunological responses, and metabolic cycles, are impacted by protein-protein interactions. These interactions have enormous implications for figuring out the origins of diseases and creating treatments. However, experimental methods for identifying PPIs are resource-intensive, time-consuming, and have limited coverage. Thus, computational techniques are essential to help and enhance activities related to protein identification. This study aims to build a deep learning network for predicting protein-protein interactions using only sequence information. Three different encoding methods are used to encode protein sequences: Binary Encoding, Autocovariance, and Position Specific Scoring Matrix. In order to predict protein-protein interactions, a convolutional Siamese neural network is employed to find complex patterns between protein sequence pairs. This network consists of two identical subnetworks with matched parameters. When applied to the human dataset, the suggested technique shows strong prediction performance with an accuracy of 84.07%, sensitivity of 92.45%, and precision of 91.45% for the model using the PSSM protein representation approach. An ensemble approach is suggested to combine the outputs from these three encoders because it is known that different encoding techniques capture various aspects of the same protein sequence. The accuracy obtained increased to 86.27% for the ensemble approach on the test set, with a sensitivity of 93.07% and a precision of 92.15%. The outcome highlights the importance of integrating several encoding methods to benefit from their complementary features and raise the accuracy of protein-protein interaction prediction. PubDate: Tue, 23 Jul 2024 00:00:00 +030
Authors:Melisa Ergün; Beyza Fırat, Gamze Tuncer, Osman Sönmez Abstract: Today, climate change, which is a threat for the whole world, especially for our country, adversely affects the structure and functioning of the ecosystem and poses a risk to our natural resources. The effect of changes in temperature, especially on the precipitation parameter, leads to a decrease in water resources. The depletion of water, which is our most important source of life, threatens our country in agricultural and socio-economic terms. In this study, Mann-Kendall and Sen’s slope trend analyses, which are nonparametric methods, were applied using monthly total precipitation, average relative humidity and average temperature data of Izmir Bolge, Kusadasi, Cesme, Odemis, Seferihisar and Selcuk stations located in Kucuk Menderes Basin, which is one of the water poor basins of our country, between 1972-2023. As a result of the study, the trend trends of meteorological data showed similar results in all stations in the analyses made by both methods. PubDate: Tue, 23 Jul 2024 00:00:00 +030