Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collection
Permanent URI for this collectionhttps://hdl.handle.net/11147/7148
Browse
7 results
Search Results
Article A Robust Possibilistic Semi-Supervised Fuzzy Clustering Algorithm With Neighborhood-Aware Feature Weighting(Springer Heidelberg, 2025) Moghaddam, Arezou Najafi; Aghazadeh, Nasser; Hashemzadeh, Mahdi; Oskouei, Amin GolzariThe Semi-Supervised Fuzzy C-Means (SSFCM) method integrates class distribution information with fuzzy logic to overcome the challenges of semi-supervised clustering methods. While the inclusion of label information in the objective function improves the quality of the clustering method, semi-supervised fuzzy techniques still encounter important limitations, including (1) sensitivity to noise and outliers, (2) uniform feature importance, (3) neglecting the influences of neighborhood in the clustering process. In this paper, an improved semi-supervised clustering algorithm is presented to address these challenges. First, the algorithm reduces the sensitivity to noise and outliers by integrating the possibilistic fuzzy C-means algorithm into the SSFCM method. Second, a dynamic feature weighting method assigns different weights to the features in each cluster, which improves the performance of the algorithm in imbalanced datasets. Third, the proposed algorithm introduces a neighborhood mechanism that incorporates the neighbor's trade-off weighting and feature weighting strategy considering a strong metric. Finally, a robust kernel metric is used to further improve the performance on complex and nonlinear datasets. Extensive experiments are conducted on several benchmark datasets to evaluate the performance of the proposed method. The results show that the proposed method outperforms the current state-of-the-art techniques. The implementation source codes of the proposed method are publicly available at https://github.com/Amin-Golzari-Oskouei/RPSSFC-NAFW.Article The Future of Urban Hierearchy and Zipf Law: Arima and Bats Forecasting(Springer Heidelberg, 2025) Duran, Hasan EnginZipf's Law is recognized as a power law which is used to identify the extent and the evolution of the urban hierarchies. The existing studies have mostly adopted a retrospective view by analysing the past patterns. However, we would like to shed a light onto the future trajectories. Therefore, the aim of this study is to investigate the future of Urban Hierarchies and Zipf's Law for the U.S. Metropolitan Statistical Areas (MSA) and the period 1969-2070. Having applied, two forecasting methods; i."ARIMA (Autoregressive Integrated Moving Average)", ii. "BATS (Exponential smoothing state space model Box-Cox transformation, ARMA errors, Trend and Seasonal components)" and the estimation of rank-size rule, we obtained crucial conclusions (Box and Jenkins in: Time series analysis: forecasting and control, Holden-Day, San Francisco, 1970; Box et al. in: Time series analysis: forecasting and control, Wiley, New Jersey, 2016; Kinney in Acc Rev 53:48-60, 1978; Hyndman et al. in R package version 8.24.0, https://cran.r-project.org/web/packages/forecast/forecast.pdf, 2025; De Livera in: Automatic forecasting with a modified exponential smoothing state space framework, Department of Econometrics & Business Statistics, Monash University (Working Papers 10/10). https://www.monash.edu/business/econometrics-and-business%20statistics/research/publications/ebs/wp10-10.pdf, 2010; De Livera et al. in: Forecasting time series with complex seasonal patterns using exponential smoothing. (Working paper 15/09), Department of Econometrics & Business Statistics, Monash University. https://robjhyndman.com/papers/ComplexSeasonality.pdf, 2010; De Livera et al. in J Am Stat Assoc 106:1513-1527, 2011). We provide evidence that the Zipf's Law is observed not to hold over the last century and, if existing conditions hold, it is not expected to be valid in the next 50 years. Pareto exponent is found significantly below the Pareto level, historically, currently and prospectively.Article Citation - WoS: 4Citation - Scopus: 5Diffusion-Based Data Augmentation Methodology for Improved Performance in Ocular Disease Diagnosis Using Retinography Images(Springer Heidelberg, 2024) Aktas, Burak; Ates, Doga Deniz; Duzyel, Okan; Gumus, AbdurrahmanDeep learning models, integral components of contemporary technological landscapes, exhibit enhanced learning capabilities with larger datasets. Traditional data augmentation techniques, while effective in generating new data, have limitations, especially in fields like ocular disease diagnosis. In response, alternative augmentation approaches, including the utilization of generative AI, have emerged. In our study, we employed a diffusion-based model (Stable Diffusion) to synthesize data by faithfully recreating crucial vascular structures in the retina, vital for detecting eye diseases by using the Ocular Disease Intelligent Recognition dataset. Our goal was to augment retinography images for ocular disease diagnosis using diffusion-based models, optimizing the outputs of the fine-tuned Stable Diffusion model, and ensuring the generated data closely resembles real-world scenarios. This strategic approach resulted in improved performance in classification models and augmentation outperformed traditional methods, exhibiting high precision rates ranging from 85% to 76.2% and recall values of 86%, and 75% for 5 classes. Beyond performance enhancement, we demonstrated that the inclusion of synthetic data, coupled with data reduction using the t-SNE method, effectively addressed dataset imbalance. As a result of synthetic data addition, notable increases of 3.4% in the precision metric and 12.8% in the recall metric were observed in the 7-class case. Strategically synthesizing data addressed underrepresented classes, creating a balanced dataset for comprehensive model learning. Surpassing performance improvements, this approach underscores synthetic data's ability to overcome the limitations of traditional methods, particularly in sensitive medical domains like ocular disease diagnosis, ensuring accurate classification. The codes of the study will be shared on GitHub in a way that benefits everyone interested: https://github.com/miralab-ai/generative-data-augmentation.Article Citation - WoS: 4Citation - Scopus: 4The Future of Regional Inequalities: an Arima Forecast(Springer Heidelberg, 2024) Duran, Hasan EnginThe existing stream of empirical literature on regional inequalities has always adopted a retrospective look by analyzing the past evolution. We depart from the main stream by adopting a future perspective: Will regional inequalities shrink over time? How will the shape of income distribution evolve? Will spatial dependency increase? In the current paper, we forecast the long-term trajectory of per capita real personal income for U.S. states using the ARIMA model. We estimate the future of disparity level (for 2050 and 2090), the shape and spatial pattern of income distribution, convergence trend and spatial dependence by the help of inequality indexes (Atkinson, Coefficient of Variation, Theil) Kernel probability density distributions, explorative maps and Moran's I test. The dataset includes 48 coterminous U.S. states over the period 1929-2022. A set of important results appeared to emerge as an outcome of the empirical analyses: First, income disparities are expected to increase over the long-term period that implies a divergence pattern. Second, the forecasted shape of the income distribution is bi-modal and polarized, therefore, pointing to a widening of the inequalities. Third, the geography of the prosperity is projected to change in a way that the geographical position of high and low-income areas will change. Fourth, spatial dependence in per capita income is expected to fade away in the future. From a political stand point, additional resources should be devoted to the states that are expected to become backward (for some states in Northeast and Southwest) in order to maintain territorial cohesion.Article Citation - WoS: 3Citation - Scopus: 4Enhancing Biogas Production From Chicken Manure Through Vacuum Stripping of Digestate(Springer Heidelberg, 2023) Sengur, Ozlem; Akgul, Deniz; Bayrakdar, Alper; Calli, BarisThe vacuum stripping's combined ammonia removal and disintegration effect on chicken manure digestate was evaluated for the first time at different pH values (8.5, 9.5, and 10.5) and temperatures (30, 50, and 70 degrees C). In this way, the potential increase in biogas production by recirculating the vacuum-stripped digestate to the anaerobic digester was determined. Experimental results showed that increasing pH and temperature significantly increase TAN removal, but pH is more effective. A significant portion of the ammonia was removed in the first 30 min. Therefore, a second set of stripping tests was performed for 30 min and at 70 degrees C and pH 10.5. After 30-min tests, a biomethane potential (BMP) assay was performed using the vacuum-stripped digestate to determine how vacuum stripping affects biomethane production. Despite having the lowest disintegration efficiency, the highest biomethane potential (56.2 +/- 29.7 mL CH4/gVS) was obtained with the digestate, which was subjected to vacuum stripping at 70 celcius without pH adjustment, and 48.7% more methane was produced than the control set. The lower residual biomethane potential in vacuum-stripped digestate at pH 9.5 and 10.5 was attributed to Na+ inhibition resulting from high NaOH consumption for pH adjustment.Article Enhanced Production of 3-Phenyllactic Acid From Novel Non-Axenic Coculture: Adaptive Evolution and Statistical Fermentation Studies(Springer Heidelberg, 2024) Meruvu, HarithaThis research pivots around screening of idoneous lactic acid bacteria (LAB) from cow milk and subjecting them to adaptive evolution experiments to aid superior growth/robustness necessary for 3-phenyllactic acid (3-PLA) production. Conventional and statistical fermentation studies were conducted at batch scale using a non-axenic coculture of three novel LAB strains: Lactiplantibacillus plantarum str. nov. plantharim, Lactobacillus delbrueckki str. nov. delharim, and Pediococcus pentasaceous str. nov. pentharim. Statistically optimized fermentation using Box Behnken technique resulted in 1225 mg/L 3-PLA production using the growth medium: cheese whey-MRS medium mixture (5:2 ratio), phenylalanine (2.69% w/v), and glucose (9.6% w/v). Statistical optimization of fermentation parameters resulted in a substantial increase (17 times higher) compared to the non-optimized fermentation conditions (72 mg/L). Monad growth kinetics of the cow milk whey (CMW) coculture were calculated and estimated as: mu(max)=0.336 h(-1), K-s=11.64 mg/mL, Y-x/s=0.835 mg/g, Y-P/S=1.66 mg/g, Y-X/P=0.112 mg/mg. The purified 3-PLA (1.93 mg/mL) showed antimicrobial activity with pathogenic bacteria like Pseudomonas aeruginosa, Escherichia coli, and Staphylococcus aureus, with a minimum inhibitory concentration of 12 mg/mL.Article Citation - WoS: 5Citation - Scopus: 5The Potential of Walnut Shells for Production of Oligosaccharides by Liquid Hot Water Treatment(Springer Heidelberg, 2023) Surek, Ece; Sabanci, Kevser; Buyukkileci, Ali OguzWalnut shell (WS), which is discarded in a large amount, is usually utilized for heating purposes; therefore, obtaining fuctional products can add value to this waste biomass. In this study, xylan was determined as the dominant carbohydrate (18.6% of dry weight) in WS. The potential applicability of liquid hot water (LHW) treatment to WS was investigated in order to solubilize hemicellulose and hydrolyze it into oligomers such as xylooligosaccharide (XOS) as a prebiotic oligosaccharide and recover solid and liquid fractions, which can be raw materials for other value-added products. LHW was applied at different temperatures (170-210 degrees C) for various times (15-120 min), and their effect was combined calculating severity factor (log R-o = 3.39-4.74). The solubilization of biomass was increased (up to 60.9%) with severity. Under optimum conditions (log R-o of 3.95, 190 degrees C-15 min), 81.5% of xylan was hydrolyzed and recovered as mainly XOS (69.8% of xylan), and also xylose and arabinose. The total oligosaccharide (XOS, arabino, gluco- and galacto-oligosaccharides) and monosaccharide (mainly xylose) concentration were 14.3 and 2 g/L, respectively, and by-products did not exceed 1.6 g/L. Moreover, 2.5 mg GAE/mL of total phenolics were obtained at those conditions, whereas that was raised to 3.4 mg GAE/mL at harsher conditions. This study presented that LHW treatment was an eco-friendly alternative method for valorization of WS through production of a liquid with high value-added compounds such as oligosaccharides and solid rich in cellulose and lignin.