Jurnal Politeknik Negeri Batam (PoliBatam)
Not a member yet
3001 research outputs found
Sort by
Implementation of the K-Nearest Neighbor Algorithm for Birth Rate Prediction
Get PDFThis study aims to predict the monthly birth rate using the K-Nearest Neighbor (KNN) regression algorithm. The dataset consists of historical data from 2010 to 2020, covering six districts and including variables such as total population, number of couples of reproductive age, family planning participation rate, and monthly birth rate as the prediction target. Data preprocessing involved handling missing values and applying Min-Max normalization. To maintain the time-series nature of the data, a chronological split was used, with 576 records from 2010 to 2018 for training and 216 records from 2019 to 2020 for testing. The model was evaluated using Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R-squared (R²). The best performance was achieved at K = 7, with MAE = 19.94, RMSE = 30.91, and R² = 0.34. Additionally, the KNN model was compared with Linear Regression and Decision Tree, where KNN outperformed both alternatives. The final model was implemented in a web-based application to facilitate demographic data management and automatic birth rate prediction per district. This system is expected to support policy planning in the fields of population control and public health.This study aims to predict the monthly birth rate using the K-Nearest Neighbor (KNN) regression algorithm. The dataset consists of historical data from 2010 to 2020, covering six districts and including variables such as total population, number of couples of reproductive age, family planning participation rate, and monthly birth rate as the prediction target. Data preprocessing involved handling missing values and applying Min-Max normalization. To maintain the time-series nature of the data, a chronological split was used, with 576 records from 2010 to 2018 for training and 216 records from 2019 to 2020 for testing. The model was evaluated using Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R-squared (R²). The best performance was achieved at K = 7, with MAE = 19.94, RMSE = 30.91, and R² = 0.34. Additionally, the KNN model was compared with Linear Regression and Decision Tree, where KNN outperformed both alternatives. The final model was implemented in a web-based application to facilitate demographic data management and automatic birth rate prediction per district. This system is expected to support policy planning in the fields of population control and public health
Stunting Risk Detection and Food Recommendation via Maternal Diagnosis Using the CF Method
Get PDFStunting in children often stems from maternal health conditions during pregnancy. This study aims to develop an intelligent rule-based IF–THEN system using the Certainty Factor method as a decision-support tool for the early detection of stunting risk factors. The detection is performed indirectly by diagnosing maternal health conditions during pregnancy. The knowledge base was constructed through interviews with obstetricians and nutritionists, encompassing 20 symptoms categorized into three primary conditions namely Chronic Energy Deficiency (CED), anemia, and preeclampsia. A total of 119 pregnant women from 11 villages in Muara Satu District participated as respondents. Implementation results revealed that among the respondents, 20 were identified with CED, 96 had anemia, and 3 exhibited signs of preeclampsia. Based on Certainty Factor (CF) calculations, the confidence distribution for CED included 2 respondents with CF <50%, 5 respondents within the 50–80% range, and 13 respondents with CF >80%. For anemia, 1 respondent had a CF value <50%, 4 fell within the 50–80% range, and 91 respondents had CF values above 80%. Meanwhile, for preeclampsia, all respondents exceeded the 50% CF threshold, with 1 respondent in the 50–80% range and 2 respondents >80%. In addition to diagnosis, the system provides tailored meal recommendations (breakfast, lunch, and dinner) based on the identified health conditions. Expert validation indicated a 90% agreement rate. However, results still require confirmation through clinical examinations and consultations to ensure medical accuracy
Support Vector Machine Classification Algorithm for Detecting DDoS Attacks on Network Traffic
Get PDFDistributed Denial of Service (DDoS) attacks represent a significant danger in network security because they can lead to extensive service interruptions. With these attacks increasingly mirroring regular traffic, smart and effective detection systems are essential. This research seeks to assess the efficacy of the Support Vector Machine (SVM) classification algorithm in identifying DDoS attacks in network traffic. The data utilized is CICIDS2017, focusing on the subset Friday-WorkingHours-Afternoon-DDos.pcap_ISCX.csv, which contains both legitimate traffic and DDoS attacks like DoS-Hulk, DoS-GoldenEye, and DDoS. The preprocessing stage included eliminating duplicates and null entries, label binary encoding, normalization through Min-Max Scaler, and feature selection applying the Chi-Square technique. The data was divided into 80% for training and 20% for testing purposes. The Radial Basis Function (RBF) kernel was utilized to train the SVM model, and hyperparameter optimization was performed with GridSearchCV. The evaluation of the model\u27s performance was conducted through accuracy, precision, recall, F1-score, confusion matrix, and visual representations including ROC and Precision-Recall Curves. The findings indicate that prior to tuning, the model reached an accuracy of 97%, which increased to 99% post-tuning, accompanied by an F1-score of 0.99. This shows that the SVM algorithm, when paired with appropriate preprocessing and optimization, is very efficient in identifying DDoS attacks within network traffic.Distributed Denial of Service (DDoS) attacks represent a significant danger in network security because they can lead to extensive service interruptions. With these attacks increasingly mirroring regular traffic, smart and effective detection systems are essential. This research seeks to assess the efficacy of the Support Vector Machine (SVM) classification algorithm in identifying DDoS attacks in network traffic. The data utilized is CICIDS2017, focusing on the subset Friday-WorkingHours-Afternoon-DDos.pcap_ISCX.csv, which contains both legitimate traffic and DDoS attacks like DoS-Hulk, DoS-GoldenEye, and DDoS. The preprocessing stage included eliminating duplicates and null entries, label binary encoding, normalization through Min-Max Scaler, and feature selection applying the Chi-Square technique. The data was divided into 80% for training and 20% for testing purposes. The Radial Basis Function (RBF) kernel was utilized to train the SVM model, and hyperparameter optimization was performed with GridSearchCV. The evaluation of the model\u27s performance was conducted through accuracy, precision, recall, F1-score, confusion matrix, and visual representations including ROC and Precision-Recall Curves. The findings indicate that prior to tuning, the model reached an accuracy of 97%, which increased to 99% post-tuning, accompanied by an F1-score of 0.99. This shows that the SVM algorithm, when paired with appropriate preprocessing and optimization, is very efficient in identifying DDoS attacks within network traffic
Real-Time Drug Classification Using YOLOv11 for Reducing Medication Errors
Get PDFAdvancements in digital imaging and machine learning have transformed healthcare, enabling innovative solutions for automated drug identification. This study develops an image-based system to classify pharmaceutical drugs, tackling errors arising from visual similarities in their shape, color, or size. Accurate drug identification is crucial for healthcare professionals and patients to access reliable information on drug composition, usage instructions, and potential side effects, enhancing safety and efficiency in medical practice. The system leverages the YOLO (You Only Look Once) algorithm, renowned for its speed and precision in object detection. A dataset comprising 5,000 drug images sourced from Kaggle was curated, with annotations and augmentation techniques such as horizontal flipping, rotation, and scaling to improve model robustness. The YOLOv11 model achieved a precision of 97.4%, a recall of 97.6%, and a mean average precision (mAP@50) of 98.4%, demonstrating high reliability in real-world scenarios. Integrated with a user-friendly Tkinter interface, the system facilitates real-time drug detection and information retrieval, streamlining access to critical data. This research underscores the YOLO algorithm’s effectiveness in delivering rapid and accurate drug classification, offering a scalable solution for healthcare applications. The system’s success highlights its potential to reduce medication errors and improve patient outcomes through precise and accessible drug identification technology
Virtual Local Area Network for Optimizing Network Performance using Simplified Form of Action Research
Get PDFVirtual Local Area Network (VLAN) is an effective solution for improving network performance and efficiency in educational environments, particularly in computer laboratories. This study aims to implement and evaluate VLAN at the Computer Laboratory of Universitas Mulia using a simplified action research approach involving a single intervention cycle. The initial diagnosis revealed that the conventional flat network topology resulted in a broad broadcast domain and reduced performance under high traffic conditions. Through VLAN segmentation, the laboratory was logically divided into five segments based on functionality. Evaluation was conducted using throughput measurement and Quality of Service (QoS) indicators, including packet loss, delay, and jitter, based on TIPHON standards. The results indicated consistent improvements in data flow and stability, with reduced packet loss under 0,3% and average delay and jitter remaining below 20 ms across all segments. In addition to technical improvements, interviews with the network administrator highlighted significant benefits in bandwidth management and traffic control during examinations. This study demonstrates that a simplified approach can be replicated in similar educational institutions with limited resources and provides a foundation for more structured and controlled network development
A Sentiment Analysis of Public Perception Toward Pets in Public Spaces Using Logistic Regression and Word Embedding
Get PDFAddressing the complex social debate over pets in public areas, this study assesses public sentiment by analyzing a dataset of YouTube comments. We employed a machine learning pipeline beginning with data collection via the YouTube API, followed by rigorous text preprocessing and SMOTE-based class balancing for the training data. For classification, a Logistic Regression model was trained on contextual features generated by Word Embeddings (Word2Vec) and optimized through hyperparameter tuning. The final model proved highly effective, yielding a test accuracy of 92.74% with F1-scores of 0.84 for the negative class and 0.95 for the positive class. Ultimately, this research establishes an effective approach to measuring public opinion on social issues in Indonesia, providing actionable insights for public space administrators and policymakers.Addressing the complex social debate over pets in public areas, this study assesses public sentiment by analyzing a dataset of YouTube comments. We employed a machine learning pipeline beginning with data collection via the YouTube API, followed by rigorous text preprocessing and SMOTE-based class balancing for the training data. For classification, a Logistic Regression model was trained on contextual features generated by Word Embeddings (Word2Vec) and optimized through hyperparameter tuning. The final model proved highly effective, yielding a test accuracy of 92.74% with F1-scores of 0.84 for the negative class and 0.95 for the positive class. Ultimately, this research establishes an effective approach to measuring public opinion on social issues in Indonesia, providing actionable insights for public space administrators and policymakers
Detection of Ripeness in Oil Palm Fresh Fruit Bunches Using the YOLO12S Algorithm on Digital Images
Get PDFIndonesia is the world\u27s largest producer of palm oil, with a production volume reaching 46.82 million tons in 2022. This industry heavily relies on the quality of Fresh Fruit Bunches (FFB) harvests, which is determined by the accuracy of ripeness at the time of harvest. Unfortunately, ripeness assessment of FFB is still conducted manually and subjectively by field workers, posing risks to both efficiency and production accuracy. Although various studies have employed YOLOv5 and YOLOv8 for fruit ripeness detection, few have explored the potential of YOLO12s in classifying FFB ripeness in a comprehensive and efficient manner. In this study, we present the application of the YOLO12s algorithm to automatically classify the ripeness levels of oil palm FFB using digital images. The YOLO12s model was trained on 14,620 FFB images categorized into four ripeness levels: unripe, under-ripe, ripe, and overripe. Evaluation results showed a precision of 93.1%, recall of 95.9%, [email protected] of 97.8%, and [email protected]–0.95 of 78.8%. The model was able to perform inference in approximately 4.7 milliseconds per image and demonstrated good generalization despite challenges related to varying lighting conditions. These results indicate that YOLO12s holds great potential to replace subjective manual methods with a more accurate, consistent, and efficient classification solution to support the harvesting process in the palm oil industry.Indonesia is the world\u27s largest producer of palm oil, with a production volume reaching 46.82 million tons in 2022. This industry heavily relies on the quality of Fresh Fruit Bunches (FFB) harvests, which is determined by the accuracy of ripeness at the time of harvest. Unfortunately, ripeness assessment of FFB is still conducted manually and subjectively by field workers, posing risks to both efficiency and production accuracy. Although various studies have employed YOLOv5 and YOLOv8 for fruit ripeness detection, few have explored the potential of YOLO12s in classifying FFB ripeness in a comprehensive and efficient manner. In this study, we present the application of the YOLO12s algorithm to automatically classify the ripeness levels of oil palm FFB using digital images. The YOLO12s model was trained on 14,620 FFB images categorized into four ripeness levels: unripe, under-ripe, ripe, and overripe. Evaluation results showed a precision of 93.1%, recall of 95.9%, [email protected] of 97.8%, and [email protected]–0.95 of 78.8%. The model was able to perform inference in approximately 4.7 milliseconds per image and demonstrated good generalization despite challenges related to varying lighting conditions. These results indicate that YOLO12s holds great potential to replace subjective manual methods with a more accurate, consistent, and efficient classification solution to support the harvesting process in the palm oil industry
A Hybrid Framework Based on YOLOv8 and Vision Transformer for Multi-Class Detection and Classification of Coffee Fruit Maturity Levels
Get PDFDetection and classification of coffee cherries based on maturity levels present a significant challenge in agricultural product processing systems, primarily due to the high visual similarity among classes within a single bunch. This study aims to develop a multi-class detection and classification system for coffee cherries by integrating YOLOv8 and Vision Transformer (ViT) as a classification enhancer. The initial detection process is conducted using YOLOv8 to identify and automatically crop coffee cherry objects from bunch images. These cropped images are then re-classified using the Vision Transformer to improve prediction accuracy. The training process was carried out with a learning rate of 0.0001, a batch size of 16, and epoch variations of 50, 100, and 150. Evaluation results demonstrate that the integration of YOLOv8 and ViT significantly improves classification accuracy compared to using YOLOv8 alone. At 100 epochs, the YOLOv8+ViT model achieved an accuracy of 89.52%, a precision of 90.43%, and a recall of 89.52%, outperforming the standalone YOLOv8 model, which only reached an accuracy of 75.44%. These results indicate that the Vision Transformer effectively enhances classification performance, particularly for visually similar coffee cherry classes. The integration of these two methods offers a promising alternative solution for improving image-based multi-class classification in agriculture and other domains involving complex visual objects
Analysis of Islamic Social Capital in Improving Welfare in MSMEs Group Financing
Get PDFThis study aims to analyze the implementation of Islamic Social Capital in improving the welfare of members of the MSMEs (Micro, Small, and Medium Enterprises) financing group under PNM Mekaar in Jaring Halus Village. The background of this research is based on the importance of Islamic values in shaping social solidarity, trust, and collective responsibility within community economic empowerment programs. This study uses a qualitative approach with a case study method, where data were collected through observation, in-depth interviews, and documentation. The results show that Islamic values such as honesty (sidq), trustworthiness (amanah), mutual assistance (ta\u27awun), and moral discipline have become social strengths that drive smooth financing, increased business turnover, and the emergence of collective aspirations towards a blessed and independent business. Furthermore, synergy with religious figures and social institutions strengthens the potential for sustainable business development. This study concludes that Islamic Social Capital serves not only as a moral foundation but also as an effective and applicable strategy for economic empowerment
THE EFFECT OF LIQUIDITY, ASSET STRUCTURE, AND COMPANY SIZE ON CAPITAL STRUCTURE WITH COMPANY ROA AS INTERVENING VARIABLES IN PLANTATION SUB SECTOR COMPANIES LISTED ON THE INDONESIA STOCK EXCHANGE
Get PDFThe aim of this research is to empirically confirm the impact of Liquidity, Asset Structure, and Firm Size on Capital Structure where Return on Assets (ROA) would serve as the mediator. Here, liquidity specifies the company\u27s capability to meet short-term obligations, asset structure and firm size specify the composition and range of the firm\u27s business that can influence financing. This research uses all listed firms in the agricultural sub-sector on the Indonesia Stock Exchange (IDX) for the period 2016-2019, with 19 observations. Parametric and non-parametric statistical techniques are employed for data analysis to examine the study variables. The steps to the analysis include the use of normality tests to confirm the appropriate distribution of data, and hypothesis testing to verify the variable relationships. Additionally, this research includes the results of the hypotheses tests and employs Path Analysis to ascertain the pathways of influence of the variables. The outcomes of this research will be in a position to provide better understanding of the determinants of the capital structure of companies operating in the agricultural sector, as well as the role of ROA as an intervening variable in such a relationship. These findings are anticipated to serve as a guide for company management in decision-making regarding funding policy and asset management.
Keywords: liquidity, return on assets, asset structure, capital structure, company size