Publication Search

This study is based on the need to develop a more effective concentrate ration for lactating dairy cows, as existing formulations in the field are greatly influenced by the availability of ingredients and varying quality. Therefore, this study focuses on optimizing concentrate in dairy cow feed rations to meet SNI standards, which include crude protein (CP), Total Digestible Nutrients (TDN), Calcium (Ca), and Phosphorus (P), with more efficient results in terms of price and nutrition. This study uses the Whale Optimization Algorithm (WOA) metaheuristic approach, which balances the exploration and exploitation processes in finding the best solution to optimization problems. This algorithm has fewer parameters than other metaheuristics such as GA, PSO, and DE. WOA runs naturally in continuous space without the need for genetic operators such as crossover and mutation. The dataset used contains types of dairy cow feed ingredients along with nutritional requirements and prices so that researchers can process the data into efficient feed concentrate that is suitable for lactating dairy cows.

Software development effort estimation is crucial as it is one of the key factors for successful software development. This research employs Random Forest to estimate software development effort. To achieve better results, the study combines the Random Forest method with Genetic Algorithm. The results show that the China dataset provides more accurate estimation compared to the Desharnais dataset, because the China dataset uses relevant feature selection for estimation.

The growing demand for energy-efficient and intelligent thermal systems has driven significant advancements in adaptive compressor design. This paper presents a comprehensive literature review on the development of AI-based compressor systems, with a specific focus on enhancing efficiency under partial-load conditions and optimizing the utilization of residual energy. Through the synthesis of five recent high-impact studies (2020–2025), we examine the application of deep reinforcement learning (DRL), hybrid evolutionary algorithms, and neural network surrogate modeling in compressor optimization. Key findings indicate that model-based DRL combined with surrogate CFD can achieve up to 8% efficiency gains at off-design conditions. Hybrid approaches integrating Genetic Algorithms (GA) with DRL reduce optimization time by 30% while improving pressure ratios. Neural network surrogates provide high-speed, real-time performance predictions with less than 1% error, enabling mass iterative design. Furthermore, intelligent load classification using radial basis function networks (RBFN) allows adaptive response to varying operating conditions with over 95% accuracy. Collectively, these methods form a framework for intelligent, self-optimizing compressor systems capable of real-time adaptation and energy recovery. The results suggest that AI-enhanced adaptive compressors represent a transformative direction for energy-sensitive sectors, including HVAC, power generation, and sustainable industry.

Information system security faces serious challenges due to increasingly complex cyber attacks. Intrusion Detection Systems (IDS) require efficient approaches to handle high-dimensional data such as the NSL-KDD dataset with 41 features. This study aims to implement the Genetic Algorithm (GA) for feature selection on the NSL-KDD dataset to improve the efficiency and accuracy of network attack detection. The method used is computational experimental research, involving data preprocessing, GA implementation for feature selection, building a classification model using Random Forest, and performance evaluation based on accuracy, precision, recall, F1-score, and computation time. The results show that GA successfully reduced features from 41 to 12 features (70.7% reduction), significantly improving computational efficiency. However, model accuracy slightly decreased from 0.4973 to 0.4951, indicating that while GA is effective for feature selection, the elimination of certain features may reduce classification capability. The implication of this study is that GA can be used as a tool to simplify intrusion detection models, but it should be combined with parameter optimization and data imbalance handling to achieve more optimal performance.  

The selection of a thesis topic that aligns with students’ interests and competencies often poses a challenge in academic environments. Inappropriate topic selection can lead to decreased motivation and delays in completing the final project. This study aims to develop a thesis topic recommendation system based on a genetic algorithm that considers students’ interests and academic abilities. The data used include grades from core courses, results of research interest questionnaires, and a list of thesis topics provided by academic supervisors. Each topic is represented as a chromosome, while the fitness function is calculated based on the level of compatibility between student attributes and topics. The selection process employs the roulette wheel method, with single-point crossover and random mutation to generate an optimal solution population. The test results show that the recommendation system based on the genetic algorithm achieves an accuracy rate of 86.7%, higher than the keyword-matching method, which only reaches 71.2%. Therefore, this approach is proven effective in assisting students to determine thesis topics that are suitable, objective, and efficient.

Toddlers are a vulnerable age group to various types of diseases due to their immune systems that are still developing. Limited utilization of medical record data and the lack of structured information regarding disease patterns in toddlers based on age and causative factors have resulted in suboptimal prevention and treatment efforts. Therefore, an approach is needed to systematically classify toddler disease data. This study aims to apply data mining techniques using the clustering method with the K-Means algorithm to group types of diseases in toddlers based on age and causative factors. The variables used in this study include toddler age, type of disease, and causative factors. The data were obtained from RSUD Dr. R. M. Djoelham Binjai and processed using MATLAB software with three clusters. The results show that the K-Means algorithm successfully groups toddler disease data into three clusters with different characteristics. The first cluster is dominated by toddlers aged 0–11 months with appendicitis caused by genetic factors. The second cluster is dominated by toddlers aged 1–3 years with diarrhea caused by environmental factors and has the largest number of members. Meanwhile, the third cluster is dominated by toddlers aged 0–11 months with sore throat caused by environmental factors. The clustering results indicate a relationship between toddler age, disease type, and causative factors, which can be used as supporting information for decision-making in the prevention and treatment of toddler diseases.

Hair loss is one of the common health problems experienced by many people and often causes psychological impacts, particularly on self-confidence. The factors contributing to hair loss are diverse, ranging from genetics, diet, and stress to lifestyle. The lack of public knowledge about these risk factors, as well as the low level of digital literacy in the use of predictive technology, makes it difficult for people to take early preventive measures. This community service activity aims to provide education and simple training on predicting hair loss risk using the Support Vector Machine (SVM) algorithm for residents of Rantau Prapat Village. The implementation methods include a pre-test to measure initial understanding, interactive counseling on hair loss risk factors, practical simulation of risk prediction using SVM based on a simple dataset, and evaluation through a post-test. The results of the activity showed a significant increase in participants’ understanding, from an average of 45.2% in the pre-test to 81.6% in the post-test, with a participant satisfaction level reaching 92%. This counseling not only improved health literacy but also introduced the practical application of artificial intelligence in the health sector.

The daily exchange of informatics over the Internet has both eased the widespread proliferation of resources to ease accessibility, availability and interoperability of accompanying devices. In addition, the recent widespread proliferation of smartphones alongside other computing devices has continued to advance features such as miniaturization, portability, data access ease, mobility, and other merits. It has also birthed adversarial attacks targeted at network infrastructures and aimed at exploiting interconnected cum shared resources. These exploits seek to compromise an unsuspecting user device cum unit. Increased susceptibility and success rate of these attacks have been traced to user's personality traits and behaviours, which renders them repeatedly vulnerable to such exploits especially those rippled across spoofed websites as malicious contents. Our study posits a stacked, transfer learning approach that seeks to classify malicious contents as explored by adversaries over a spoofed, phishing websites. Our stacked approach explores 3-base classifiers namely Cultural Genetic Algorithm, Random Forest, and Korhonen Modular Neural Network – whose output is utilized as input for XGBoost meta-learner. A major challenge with learning scheme(s) is the flexibility with the selection of appropriate features for estimation, and the imbalanced nature of the explored dataset for which the target class often lags behind. Our study resolved dataset imbalance challenge using the SMOTE-Tomek mode; while, the selected predictors was resolved using the relief rank feature selection. Results shows that our hybrid yields F1 0.995, Accuracy 0.997, Recall 0.998, Precision 1.000, AUC-ROC 0.997, and Specificity 1.000 – to accurately classify all 2,764 cases of its held-out test dataset. Results affirm that it outperformed bench-mark ensembles. Result shows the proposed model explored UCI Phishing Website dataset, and effectively classified phishing (cues and lures) contents on websites.

The integration of Internet of Things (IoT) with cloud computing has revolutionized healthcare systems, offering scalable and real-time patient monitoring. However, optimizing response times and energy consumption remains crucial for efficient healthcare delivery. This research evaluates various algorithmic approaches for workload migration and resource management within IoT cloud-based healthcare systems. The performance of the implemented algorithm in this research, Hybrid Dynamic Programming and Long Short-Term Memory (Hybrid DP+LSTM), was analyzed against other six key algorithms, namely Gradient Optimization with Back Propagation to Input (GOBI), Deep Reinforcement Learning (DRL), improved GOBI (GOBI2), Predictive Offloading for Network Devices (POND), Mixed Integer Linear Programming (MILP), and Genetic Algorithm (GA) based on their average response time and energy consumption. Hybrid DP+LSTM achieves the lowest response time (82.91ms) with an energy consumption of 2,835,048 joules per container. The outcome of the analysis showed that Hybrid DP+LSTM have significant response times improvement, with percentage increases of 89.3%, 79.0%, 83.8%, 97.0%, 99.8%, and 99.94% against GOBI, GOBI2, DRL, POND, MILP, and GA, respectively. In terms of energy consumption, Hybrid DP+LSTM outperforms other approaches, with GOBI2 (3,664,337 joules) consuming 29.3% more energy, DRL (2,973,238 joules) consuming 4.9% more, GOBI (4,463,010 joules) consuming 57.4% more, POND (3,310,966 joules) consuming 16.8% more, MILP (3,005,498 joules) consuming 6.0% more, and the GA (3,959,935 joules) consuming 39.7% more. The result of ablation of the Hybrid DP+LSTM model achieves a 47.05% improvement over DP-only (156.57ms) and a 70.64% improvement over LSTM-only (282.41ms) in response time. On the energy efficiency side, Hybrid DP+LSTM shows 22.80% improvement over LSTM-only (3,671,51 joules), but 7.34% underperformance compared to DP-only (2,640,93). These research findings indicate that the Hybrid DP+LSTM technique provides the best trade-off between response time and energy efficiency. Future research should further explore hybrid approaches to optimize these metrics in IoT cloud-based healthcare systems.

Shadow puppet craft training is a strategic intervention in preserving cultural heritage and strengthening the creative economy sector in Indonesia. To ensure the effectiveness and efficiency of training, a planning approach is needed that is not only conventional, but also based on quantitative analysis and intelligent systems. This community service proposes a training planning strategy using an interdisciplinary approach involving Operation Research, Design of Experiment (DoE), Simulation, Metaheuristic Algorithms, and Data Mining. This study begins with the identification of key training variables, such as duration, number of participants, initial competency level, teaching materials, and instructor resources. Through the DoE approach, various combinations of variables are systematically tested to identify the optimal training design. Next, Simulation is used to model the dynamics of training implementation and evaluate implementation scenarios. To predict training needs and participant behavior, Data Mining techniques are applied to historical data of arts community training. In the final stage, Metaheuristic algorithms such as Genetic Algorithm and Simulated Annealing are used to solve complex and large-scale scheduling and resource allocation problems. The results of the integration of these approaches show an increase in training efficiency of up to 27% as well as increased participant satisfaction and the quality of work results. This activity demonstrates that applying a quantitative, data-driven approach to traditional crafts training planning can provide significant added value. This model can be replicated in other training programs based on local wisdom and other creative industry sectors.

This research explores the effectiveness of heuristic techniques for solving combinatorial optimization problems, with a particular focus on the Traveling Salesman Problem (TSP). Combinatorial optimization is a critical area of study, especially in fields like computer science, engineering, and economics, where finding optimal solutions from a finite set of possibilities is crucial. However, the NP-hard nature of many combinatorial problems, such as the TSP, makes traditional exact methods like Branch-and-Bound and Dynamic Programming computationally expensive and inefficient for larger problem sizes. The primary objective of this research is to evaluate the performance of heuristic methods, including Simulated Annealing (SA), Genetic Algorithms (GA), and Iterative Computation techniques, such as Tabu Search (TS) and Particle Swarm Optimization (PSO). These methods are tested for their ability to provide approximate solutions efficiently. The findings reveal that while ACO provided the best solution quality, it had the longest runtime. TS was the fastest, though with slightly lower solution quality. SA and GA demonstrated a balance between solution quality and computational efficiency, but their performance heavily depended on parameter tuning. The hybridization of SA and GA showed potential for improving solution quality but introduced additional complexity. The research concludes that heuristic methods, especially when combined, offer viable solutions for large-scale combinatorial optimization problems, though the trade-off between solution quality and computational time must be considered when selecting an algorithm.

Penelitian ini bertujuan untuk menganalisis optimasi parameter Photovoltaic (PV) serta daya keluaran dari pembangkit listrik tenaga surya berbasis Genetic Algorithm. Penelitian ini menggunakan metode pengumpulan data melalui studi literatur, simulasi perangkat lunak, serta pengujian model matematis. Penelitian ini menggunakan pendekatan analisis kuantitatif dan komputasional yang berfokus pada pencarian nilai optimal dari parameter-parameter utama sistem PV, seperti tegangan, arus, dan resistansi. Hasil penelitian menunjukkan bahwa dalam proses optimasi sistem PV, tahapan yang dilakukan meliputi pemodelan karakteristik PV, penyusunan fungsi objektif untuk memaksimalkan daya, penerapan algoritma genetika dalam mencari nilai parameter optimal, serta analisis hasil optimasi. Hasil penelitian memperlihatkan bahwa algoritma genetika mampu meningkatkan efisiensi daya keluaran PV dibandingkan metode konvensional, serta menghasilkan parameter yang lebih sesuai terhadap kondisi iradiasi tertentu. Namun, dalam implementasi metode ini juga ditemukan beberapa kendala, seperti kompleksitas perhitungan dan kebutuhan waktu komputasi yang relatif tinggi. Selain itu, diperlukan penyesuaian lebih lanjut untuk penerapan algoritma ini dalam skala sistem pembangkit yang lebih besar.

Pendistribusian bantuan yang efisien menjadi tantangan utama dalam proses penyaluran. Penelitian ini bertujuan untuk merancang sistem cerdas yang mampu memetakan dan menyalurkan bantuan secara optimal dengan pendekatan algoritmik. Metodologi yang digunakan meliputi pemodelan sistem dengan menggunakan Unified Modeling Language (UML), perancangan struktur basis data relasional, dan perancangan algoritma penyaluran berdasarkan kriteria prioritas dan efisiensi logistik. UML digunakan untuk menggambarkan arsitektur sistem secara visual, meliputi use case, class, dan diagram aktivitas. Perancangan basis data dilakukan untuk memastikan integritas data dan kemudahan pengelolaan informasi bantuan, lokasi, dan kebutuhan penerima. Algoritma yang dikembangkan memanfaatkan pendekatan heuristik untuk menentukan rute penyaluran dan prioritas penerima berdasarkan parameter lokasi geografis. Hasil penelitian ini berupa prototipe model konseptual yang dapat digunakan sebagai dasar pengembangan sistem cerdas berbasis teknologi untuk mendukung proses penyaluran bantuan yang adaptif dan responsif.

This study proposes a hybrid approach combining Frequent Itemset Mining (FIM) and Algorithms and Genetic Algorithm for product layout optimization with a case study at PT. MPI Pharmacy. The FIM Algorithms is employed to extract association rules from 1,000 beauty product sales transactions, while the Genetic Algorithm is utilized to perform product placement based on these rules generated, with storage space constraints. Implementation results demonstrate that this hybrid approach successfully identifies 18 key association rules (support >15, confidence >80%) and proposes an optimal layout configuration model that reduces customer travel distance by 25 compared to conventional layouts used by MPI Pharmacy. The Genetic Algorithm solves complex rule-based optimization problems for product placement, which are limited by traditional market basket analysis (MBA) approaches that rely solely on association rules. This hybrid sistem not only improves pharmacy operational efficiency at PT MPI (reducing service time by 18) but also increases cross-selling opportunities by 22. Hence, inventory operations management impproved efficiently. The research findings contribute to the field of retail space optimization by effectively integrating association rule mining and evolutionary computation.  

This paper explores the application of quantum-inspired optimization algorithms in the training of large-scale Graph Neural Networks (GNNs) within distributed cloud-edge environments. GNNs have gained significant attention due to their ability to model complex relationships in graph-structured data, yet their training presents challenges such as high computational demand, inefficient resource allocation, and slow convergence, especially for large datasets. Traditional meta-heuristic algorithms, while useful, often face scalability and performance issues when applied to such large-scale tasks. To address these challenges, we propose a quantum-inspired meta-heuristic algorithm that leverages quantum principles, such as superposition and entanglement, to enhance optimization processes. The algorithm was integrated into a hybrid cloud-edge system, where computational tasks are dynamically distributed between edge nodes and the cloud, optimizing resource utilization and reducing latency. Our experimental results demonstrate significant improvements in training speed, resource efficiency, and convergence rate when compared to traditional optimization methods such as Genetic Algorithms and Simulated Annealing. The quantum-inspired algorithm not only accelerates the training process but also reduces memory usage, making it well-suited for large-scale GNN applications. Furthermore, the system's scalability was enhanced by the hybrid cloud-edge architecture, which balances computational load and enables real-time data processing. The findings suggest that quantum-inspired optimization algorithms can significantly improve the training of GNNs in distributed systems, opening new avenues for real-time applications in areas such as social network analysis, anomaly detection, and recommendation systems. Future work will focus on refining these algorithms to handle even larger datasets and more complex GNN architectures, with potential integration into edge devices for enhanced real-time decision-making.

Efficient resource allocation is a major challenge in various sectors, especially when faced with limitations in quantity, time, and cost. This study aims to examine the application of linear programming as an optimization method in solving assignment problems, where a number of resources must be optimally allocated to a number of specific tasks. Through a literature study approach, this study examines various relevant previous study results, especially in the Indonesian context. The results of the study indicate that linear programming is able to improve operational efficiency, reduce costs, and produce a more balanced task distribution. However, this model has limitations in dealing with non-linear conditions and data uncertainty, so integration with other methods such as fuzzy logic or genetic algorithms is needed. This study is expected to broaden the understanding of the benefits of linear programming and encourage its wider application in quantitative-based decision making.

Oil palm is a key commodity supporting Indonesia’s economy through exports and employment. The industry’s success depends heavily on the selection of superior seedlings, which determine productivity, crop quality, and resistance to pests and diseases. Manual selection, however, often leads to subjectivity and inconsistency due to limited human resources and genetic variation. To address this, the study applies the Naïve Bayes algorithm for classifying oil palm seedlings based on seven variables: height, stem diameter, number of leaves, leaf color, disease resistance, root growth, and fruit yield. Using an explanatory quantitative method, the study follows seven stages: identifying problems, literature review, collecting 1,000 data entries from PT Intitama Berlian Perkebunan, data pre-processing, system modeling (UML), algorithm implementation, and evaluation using a confusion matrix and black box testing. Data was split into 80% training and 20% testing. The Naïve Bayes-based classification achieved 95% accuracy and perfect recall (1.00) for the superior seedling class. However, its performance on the minority class (non-superior seedlings) was weaker due to dataset imbalance. Black box testing verified all system functions worked correctly, enabling effective and efficient use by administrators. The study concludes that Naïve Bayes improves objectivity, efficiency, and accuracy in seedling selection. Nonetheless, attention is needed on data balancing and optimization to maintain consistent performance across classes. This system shows strong potential as a decision-support tool in plantations and promotes digital transformation in agricultural processes.

The interaction between genetic and environmental factors plays a crucial role in determining phenotypic traits in organisms. This study aims to analyze these interactions using computational approaches, including statistical models and machine learning algorithms. The data used include genetic factors (genotypes) and simulated environmental factors. Results indicate that machine learning models such as Random Forest can detect interaction patterns with high accuracy, as demonstrated by significant R² values. Additionally, heatmap visualizations provide deeper insights into the non-linear effects of genetic-environment interactions. This study highlights the potential of computational methods in exploring complex interactions, with broad applications in health, agriculture, and biotechnology.

Task scheduling in cloud computing environments is a crucial aspect in optimizing resource allocation and improving system efficiency. This research aims to analyze trends in task scheduling algorithms in cloud computing using a Systematic Literature Review (SLR) approach on various scientific publications published between 2018 and 2025. The results of the study show that Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), and Genetic Algorithm (GA) algorithms are the most commonly used methods in solving task scheduling problems. PSO stands out as an effective algorithm due to its ability to find global optimal solutions, handle non-linear and multimodal problems, and its efficiency in managing computational resources. Additionally, various studies have shown that optimization of scheduling algorithms can be achieved through a combination or modification of existing methods to improve system performance. This study provides in-depth insights into the development of scheduling algorithms in cloud computing and opens up opportunities for further research in developing more innovative and adaptive approaches.

Natural resource management involves complex decision-making processes that often result in non-linear optimization problems. This study explores the application of genetic algorithms (GA) and particle swarm optimization (PSO) to manage resources like water and forest reserves more efficiently. We compare the effectiveness of these algorithms in achieving sustainable utilization while minimizing environmental impact. The results show that GA outperforms PSO in forest management scenarios, while PSO is more suitable for water resource distribution.