Kemal Eskioglu

Kemal Eskioglu

Auto-generated profile, based on publicly available information

Kemal Eskioglu

Akdeniz Universitesi

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
2 August 2024 in Mediterranean Agricultural Sciences
Verified
Journal Article
Open access
PDF file
Berkant İsmail YILDIZBerkant İsmail YILDIZ
Berkant İsmail YILDIZ
KEKemal Eskioğlu
Kemal Eskioğlu
KKKemal Karabağ
Kemal Karabağ
Türkiye, with its rich flora diversity, holds a significant share in global honey production. However, honey bee populations, essential for agricultural ecosystems, face multifaceted threats such as climate change, habitat degradation, diseases, parasites, and exposure to pesticides. Alongside the increasing global food demand driven by population growth, there is a pressing need for a substantial increase in honey production. In this context, advances in machine learning algorithms offer tools to predict future food needs and production levels. The objective of this work is to develop a predictive model using machine learning techniques to predict Türkiye's honey output in the next years. To achieve this goal, a range of machine learning algorithms including K-Nearest Neighbor, Random Forest, Linear Regression, and Gaussian Naive Bayes were employed. Following investigations, Linear Regression emerged as the most effective method for predicting honey production levels (R2= 0.97).

1 April 2025 in Brazilian Journal of Poultry Science
Verified
Journal Article
Open access
Berkant İsmail YILDIZBerkant İsmail YILDIZ
Berkant İsmail YILDIZ
KEK Eskioğlu
K Eskioğlu
Demir OzdemirDemir Ozdemir
Demir Ozdemir
MAMUSTAFA - AKŞİT
MUSTAFA - AKŞİT
This study evaluates the effectiveness of machine learning algorithms in predicting quail egg quality based on nine key parameters, including egg weight, egg width, egg length, yolk height, yolk width, yolk weight, albumen height, albumen width, and albumen length. A dataset comprising 350 eggs from 18-week-old Japanese quails was analyzed using Logistic Regression, Naive Bayes, Support Vector Machines, k-Nearest Neighbors, Random Forest, and Gradient Boosting. The findings revealed that models combining internal and external quality parameters achieved significantly higher accuracy compared to models based solely on external attributes. Notably, Random Forest and Gradient Boosting algorithms achieved accuracies exceeding 97%, while predictions based only on external parameters exhibited lower accuracy but presented a promising starting point for non-invasive evaluations. This study strongly highlights the applicability and flexibility of machine learning in evaluating quail egg quality. The ability of algorithms to integrate various quality parameters and analyze complex relationships provides both rapid and scalable solutions. These findings demonstrate that machine learning technologies have the potential to drive innovative approaches in the poultry industry and inspire future research focusing on larger datasets and additional parameters to further enhance accuracy.

16 December 2025 in Turkish Journal of Agriculture - Food Science and Technology
Verified
Journal Article
Closed access
KEKemal Eskioğlu
Kemal Eskioğlu
Demir OzdemirDemir Ozdemir
Demir Ozdemir
Poultry farming has become a vital component of human nutrition worldwide, growing alongside industrial production. Despite its scale reaching millions of tons annually infectious diseases remain a major challenge. Each year, millions of chickens reportedly die due to these outbreaks, causing significant economic losses and disrupting flock health. Traditional methods often fall short in early detection and control, making timely intervention difficult. Recent technological advancements have opened new possibilities in poultry health management. Artificial Intelligence (AI), in particular, has emerged as a promising solution in a sector that is otherwise hard to monitor effectively. AI technologies, mainly categorized into Machine Learning (ML) and Deep Learning (DL), enable the analysis and interpretation of large, complex datasets. Using AI-based models, researchers can analyze both collective and individual behaviors, along with biological data, to identify early signs of disease, often before visible symptoms occur. Studies have shown that ML algorithms such as Support Vector Machine (SVM) and Random Forest can achieve accuracy rates exceeding 90%. Meanwhile, DL models like Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks have demonstrated predictive accuracy as high as 94–98%. These systems help producers identify not just diseases but also potential pathogens at early stages, significantly minimizing animal losses and enhancing biosecurity. However, for widespread adoption, challenges like data quality, implementation cost, and farmer training need to be addressed. In conclusion, AI-powered disease prediction tools represent an innovative and impactful development in the poultry industry. By improving early detection, reducing economic losses, and enhancing operational efficiency, they are expected to play a crucial role in shaping the future of sustainable, technology-driven livestock production.

20 August 2025 in Mediterranean Agricultural Sciences
Verified
Journal Article
Open access
PDF file
KEKemal Eskioglu
Kemal Eskioglu
Berkant İsmail YILDIZBerkant İsmail YILDIZ
Berkant İsmail YILDIZ
Demir OzdemirDemir Ozdemir
Demir Ozdemir
Chicken meat production plays a vital role in the global food industry by providing a cost-effective and sustainable protein source for a rapidly growing population. Given strategic importance, accurately estimating production levels is essential for improving operational efficiency, optimizing resource use, and responding to market demand. In recent years, data-oriented methods have become integral to modern agriculture, with machine learning models emerging as powerful tools for modeling agricultural outputs. This study aims to develop and compare predictive models for chicken meat production in Türkiye using four machine learning algorithms: Linear Regression (LR), Random Forest (RF), k-Nearest Neighbors (k-NN), and Extreme Gradient Boosting (XGBoost). The models were trained on a comprehensive dataset spanning 62 years (1961–2022), incorporating meteorological, agricultural, and economic variables. Key predictors such as feed use, weight gain, and environmental factors were included. The dataset was carefully prepared to ensure robust model training and validation. Model performance was evaluated using multiple metrics, including the coefficient of determination (R²), MAE, MSE, and RMSE. Results indicated that the Linear Regression model achieved the highest R² value and the lowest error rates among all algorithms. These findings underscore the potential of AI-based approaches to enhance decision-making and resource management in the poultry sector. With further integration of diverse data sources and advanced learning techniques, this framework can contribute to the development of more efficient, adaptive, and sustainable poultry production systems.

6 December 2024 in Mediterranean Agricultural Sciences
Verified
Journal Article
Open access
PDF file
KEKemal Eskioglu
Kemal Eskioglu
Berkant İsmail YILDIZBerkant İsmail YILDIZ
Berkant İsmail YILDIZ
Demir OzdemirDemir Ozdemir
Demir Ozdemir
Multiple Sequence Alignment (MSA) is an essential method in bioinformatics for detecting conserved sequence regions and deducing evolutionary relationships. However, performance variability exists among MSA tools, and different tools yield varying results depending on the dataset. This study conducts a comparative evaluation of four widely used MSA tools: ClustalW, Clustal Omega, MUSCLE, and MAFFT. The alignment quality and processing efficiency of these tools were assessed using 40 randomly selected Gallus gallus cytochrome c oxidase subunit 1 (COX1) DNA sequences. The findings offer valuable insights into the specific contexts in which these tools may be most effective. MAFFT demonstrated a notable advantage in processing speed, while Clustal Omega and MAFFT excelled in Column Score (CS). For Total Consensus (TC) score, ClustalW and MUSCLE showed superior performance, and Clustal Omega exhibited the highest performance based on Root Mean Square Deviation (RMSD) values. No significant difference was observed between the tools in terms of the Sum-of-Pairs (SP) score. This study serves as a valuable resource for researchers seeking to optimize the use of MSA tools for their specific applications.

16 December 2025 in Mediterranean Agricultural Sciences
Verified
Journal Article
Open access
Berkant İsmail YILDIZBerkant İsmail YILDIZ
Berkant İsmail YILDIZ
KEKemal Eskioglu
Kemal Eskioglu
Demir OzdemirDemir Ozdemir
Demir Ozdemir
This study presents the development of a novel Python-based pipeline for the identification of optimal single-guide RNA (sgRNA) target sites in CRISPR-Cas9 gene knockout applications. The pipeline offers a comprehensive analysis framework for targeting the Myostatin gene in Gallus gallus, assessing both the efficiency and specificity of potential sgRNA sites. By integrating state-of-the-art bioinformatics tools and databases during the design phase, the pipeline has been rigorously tested across various genetic models, demonstrating superior performance relative to existing software. Notably, our pipeline predicted a maximum efficiency of 100% in targeting the Myostatin gene, outperforming the CHOPCHOP and E-CRISP tools by identifying novel, high-potential target sites. Additionally, the pipeline’s user-friendly interface and interactive visualization capabilities enhance its accessibility, making it an invaluable resource for researchers and biotechnological applications in CRISPR-Cas9 genome editing. This work aims to advance gene editing precision, streamline workflows, and establish a new benchmark for genetic engineering technologies.