Publication Search

The development of artificial intelligence deepfake technology has opened up new opportunities in various fields to help speed up human work. However, on the other hand, this technology can also be misused to commit crimes. This research is a Normative Juridical research with a statutory approach and a conceptual approach, and examines the sources of legal materials according to the main problem, and uses prescriptive analysis techniques.The results of the study show: (1) The activity of using artificial intelligence deepfake technology that can cause cybercrime occurs due to attacks on the system, namely AI botnet attacks that have been infected by malicious software and Generative Adversarial Network attacks that have artificial neural networks that can produce data that is similar to the original data so that it is used as a means of committing crimes, and (2) Criminal law regulations in Indonesia regarding the misuse of artificial intelligence deepfake technology in committing cybercrime have not been regulated comprehensively, so that currently it is necessary to establish clear legal regulations in order to provide legal protection for every community.

This research investigates school scope classification using Deep Neural Networks (DNN), focusing on students living environments and educational opportunities. By addressing the interplay of socioeconomic and educational factors, the study aims to develop an analytical framework for understanding how environmental contexts shape academic trajectories. The research provides a nuanced understanding of the importance of features in educational classification by developing DNN models based on Spearman's Rank Correlation Coefficient (SRCC). The methodology employs machine learning techniques, integrating data wrangling, exploratory analysis, and multiple DNN models with K-fold cross-validation. The study analyzes 677 student records from two schools. The research examined multiple model configurations. Results show that the 'All Data' model achieved 83.08% accuracy, the 'Top 5' model 81.54%, and the 'Non-Top 5' model 79.23%. The SRCC-based approach revealed that while top correlated features are important, additional variables significantly contribute to model performance. The study highlights the profound impact of family background, social environment, and educational contexts on school selection. Furthermore, it demonstrates DNN's capability to uncover intricate, non-linear relationships, offering actionable insights for policymakers to leverage machine learning's potential in developing targeted educational strategies.

Facial Emotion Recognition (FER) is a key technology for identifying emotions based on facial expressions, with applications in human-computer interaction, mental health monitoring, and customer analysis. This study presents the development of a real-time emotion recognition system using Convolutional Neural Networks (CNNs) and OpenCV, addressing challenges such as varying lighting and facial occlusions. The system, trained on the FER2013 dataset, achieved 85% accuracy in emotion classification, demonstrating high performance in detecting happiness, sadness, and surprise. The results highlight the system's effectiveness in real-time applications, offering potential for use in mental health and customer behavior analysis.

This study presents the development and evaluation of an automatic passenger counting system for public buses using the YOLOv8 algorithm based on Convolutional Neural Networks (CNN). Accurate passenger counting plays a crucial role in optimizing public transportation operations, as it enables effective capacity management, reduces operational costs, and improves overall passenger comfort. Conventional manual counting methods are often inefficient, time-consuming, and prone to human error, particularly in high-density urban transportation environments. Therefore, an automated and intelligent solution is required to support real-time monitoring and operational decision-making. The proposed system employs deep learning-based object detection to identify and count passengers from video streams captured by cameras installed inside buses. Two camera positions, namely front and rear views, were evaluated to assess system performance under different visual conditions. The experimental results show that the system achieves high detection accuracy in the front camera view, with a confidence score of 0.82, indicating reliable performance in scenarios with minimal object occlusion. In contrast, the rear camera view demonstrates slightly lower accuracy, with a confidence score of 0.76, mainly due to increased object overlap and variations in lighting conditions. These findings emphasize the importance of appropriate camera placement and environmental consideration in improving detection reliability. In addition, the implementation of the proposed system enables real-time monitoring of passenger flow, which supports dynamic scheduling, demand-based route planning, and efficient fleet management. Accurate passenger data allows transportation operators to optimize service allocation, reduce congestion, and enhance overall service quality. Overall, this study contributes to the development of intelligent transportation systems by demonstrating the practical applicability of deep learning-based passenger counting solutions. The proposed approach offers strong potential for real-world deployment in smart city environments, supporting the creation of more sustainable, efficient, and passenger-oriented public transportation services.

This study discusses the implementation of computer vision technology for face detection in photos using two sample images with variations in lighting and face pose. The developed system combines the Viola-Jones algorithm and Convolutional Neural Networks (CNN) to enhance resilience against lighting and face orientation variations. Experimental results show high accuracy even with only two sample images. This research also develops preprocessing techniques to handle extreme lighting conditions and demonstrates efficient implementation using Python and OpenCV.  

Dangeu dengue fever or what we often call dengue fever is a disease transmitted by the Aedes aegypti mosquito and caused by the dengue virus. This disease can potentially cause serious complications if it does not receive proper treatment. In this research, the author uses the application of artificial neural networks with the Hebb rule approach to predict the risk level of dengue fever. Predictions are made based on factors such as weather conditions, population density and historical case data that influence this disease. The Hebb rule is used in this research because of its ability to strengthen connections between neurons based on the input patterns they receive, so it is hoped that it can produce more accurate predictions. Test results show that this method has a fairly high level of accuracy in predicting the pattern of dengue fever cases in an area. This research indicates that the application of artificial neural networks with the Hebb rule can be an effective tool for related parties in taking preventive measures to minimize the number of dengue cases in the future.    

The selection of agricultural and plantation products often relies on human perception of fruit color. Manual identification through visual observation has several drawbacks, such as time consumption, fatigue, and varying perceptions of quality. Digital image processing technology enables automatic sorting of products. This study applies the Perceptron learning method to identify tomato ripeness. Tomato images are captured using a webcam, analyzed through color histograms, and identified using artificial neural networks. The identification success rate reaches 43.33%, with outputs categorized as Unripe (10%), Half-Ripe (6.66%), and Ripe (26.66%).

The use of Artificial Neural Networks (JST) for weather prediction is one of the innovative approaches in climate data analysis. This study aims to apply JST in predicting weather, especially rainfall and the number of rainy days in the North Sumatra region. Historical weather data obtained from BMKG Region I for 2022-2023 is used as input to train the JST model. With a training process that involves processing rainfall data, this model is expected to provide accurate predictions regarding weather patterns. The results of this research can help in agricultural sector planning, disaster risk mitigation, and natural resource management. JST has proven to be effective in identifying dynamic and complex weather patterns, so it has the potential to be used in long-term weather prediction.

Motivation has an important role in the teaching and learning process for both teachers and students. For teachers, knowing students' learning motivation is very necessary. maintain and increase students' enthusiasm for learning. For students, learning motivation can foster enthusiasm for learning so that students are encouraged to carry out learning actions. Students carry out learning activities happily because they are driven by motivation. Currently, many students are less motivated to study. Backpropagation is a supervised learning algorithm and is usually used by perceptrons with many layers to change the weights connected to neurons in the hidden layer. Based on the learning rate and maximum epoch values, artificial neural networks using the backpropagation method can predict the level of student learning motivation with convergent results or the target error is achieved with an epoch of 11 iterations and a training process time (time) of 0.00.08 seconds. From the student learning motivation criteria data which is used as training data, the training targets can be identified. Yes and no input which is transformed into 0 and 1 can predict the level of student learning motivation with low, medium and high student motivation targets with reslt testing 80%.

The rapid development of technology in the globalization era has significantly impacted various aspects of life, including the healthcare sector. RSU Bidadari Binjai, as a healthcare provider, faces challenges in diagnosing and preventing Gastroesophageal Reflux Disease (GERD), a condition with high prevalence and serious complications such as Barrett’s esophagus and esophageal cancer. Therefore, a predictive system capable of early detection is needed to ensure quicker and more effective medical intervention. This research develops a computer-based predictive system using the backpropagation method in artificial neural networks to assist in diagnosing GERD by processing patient symptom data. The system's test results show an accuracy rate of 100% in predicting GERD complications based on the given symptoms, thus supporting more timely and accurate medical interventions.    

Every company is required to plan the need for goods as effectively as possible in order to maximize profits. Bintang Makmur Building Shop is a building shop that provides building materials, especially cement. Cement is one of the basic materials for buildings. The need for cement has recently continued to increase due to the large number of developments, both housing projects and road construction. In addition to the increasing demand for cement, cement prices also experienced price volatility which tended to fluctuate. This is done so that there is no stockpiling or even a shortage of cement. With prices that tend to go up and down if there is too much stock, it will cause losses if there is a price decrease. Vice versa if there is a shortage of cement stock, it can cause disappointment to customers. To deal with the above, it is necessary to build a prediction system that can predict cement needs in prosperous shops. The system that will be built uses an Artificial Neural Network (Artificial Neural Network) which is part of the science of artificial intelligence which has been widely used to solve various kinds of problems related to prediction or forecasting by utilizing the Backpropagation Method. The system is designed with the MATLAB programming application. From the results of the research that has been carried out, it was found that the total demand for Andalas cement for January of the following year is 0.2532 or 2532, thus the predicted total demand for Andalas cement is 2532 sacks.

Accurate rainfall prediction is needed to improve the performance of land that always uses rainfall data. Data mining or often called knowledge discovery in databases (KDD) is an activity that includes collecting, using historical data to find regularities, patterns or relationships in large data. In predicting rainfall, there are several conditions that can be observed as reference data to predict rainfall, namely wind speed, temperature, and air humidity. In this research, a backpropagation artificial neural network prediction method is developed that can be used in predicting future rainfall. The backpropogation artificial neural network method that was built produced an accuracy value of 95.36%, a precision value of 90.50%, a recall value of 97.50% and an f-measure value of 92.00%

The rapid growth of healthcare services has significantly increased the generation of medical waste, posing serious risks to environmental and public health, particularly through water quality contamination. This study examines global trends in medical waste management and proposes a hybrid analytical framework combining bibliometric analysis using SCIMAT and predictive modeling based on artificial neural networks implemented in Orange. Bibliometric mapping was employed to identify dominant research themes, temporal evolution, and knowledge gaps in medical waste and water contamination studies from 2000 to 2023. Subsequently, a neural network model was developed to predict potential water quality deterioration associated with mismanaged medical waste, using simulated environmental indicators. The results reveal a strong research focus on incineration, infection control, and hazardous waste, while predictive modeling of water contamination remains underexplored. The proposed hybrid approach demonstrates high predictive accuracy and offers a robust decision-support tool for environmental health policy. This study contributes methodologically and substantively to sustainable medical waste management and water resource protection.

Industrial IoT (IIoT) networks, critical for automation and smart manufacturing, are susceptible to faults due to their complexity and the large number of connected devices. This paper introduces a deep learning-based approach for early fault detection in IIoT networks. By leveraging recurrent neural networks (RNNs) and convolutional neural networks (CNNs), the system effectively identifies anomalies in real-time, helping to reduce system downtime and enhance operational efficiency in industrial settings.

This study analyzes the use of deep learning, primarily Convolutional Neural Networks (CNN), to categorize various types of citrus fruits. The study attempts to create an automated system that can accurately categorize citrus fruit kinds using image processing techniques. The collection contains 40 photos of four different citrus fruit types: pomelo, mandarin orange, kaffir lime, and lime. The methodology entails gathering photos, preprocessing them to improve quality, and then training a CNN model to classify the fruit varieties. The results show a high accuracy rate of 95% in classifying fruit types, demonstrating that the CNN model is effective for this task. The findings indicate that increasing the dataset and including other fruit species could significantly boost the system's accuracy.

This study proposes an approach for human fall classification utilizing a combination of Weighted Moving Average (WMA) and Convolutional Neural Networks (CNN) on the SisFall dataset. Falls among elderly individuals pose a significant public health concern, necessitating effective automated detection systems for timely intervention and assistance. The SisFall dataset, comprising accelerometer data collected during simulated falls and activities of daily living, serves as the basis for training and evaluating the proposed classification system. The proposed method begins by preprocessing accelerometer data using a WMA technique to enhance signal quality and reduce noise. Subsequently, the preprocessed data are fed into a CNN architecture optimized for feature extraction and fall classification. The CNN leverages its ability to automatically learn discriminative features from raw sensor data, enabling robust and accurate classification of fall and non-fall events. Experimental results demonstrate the efficacy of the proposed approach in accurately distinguishing between fall and non-fall activities, achieving high classification performance metrics such as accuracy, precision, recall, and F1-score. Comparative analysis with existing methods showcases the WMA-CNN hybrid approach's superiority in classification accuracy and robustness. Overall, the proposed methodology presents a promising framework for real-time human fall classification using sensor data, offering potential applications in wearable devices, ambient assisted living systems, and healthcare monitoring technologies to enhance safety and well-being among elderly individuals.

Leukemia, a global health challenge characterized by malignant blood cell proliferation, demands innovative diagnostic techniques due to its increasing incidence. Among leukemia types, Acute Lymphoblastic Leukemia (ALL) emerges as a particularly aggressive form affecting diverse age groups. This study proposes an advanced mechanized system utilizing Deep Neural Networks for detecting ALL blast cells in microscopic blood smear images. Achieving a remarkable accuracy of 97% using MobileNetV2, our system demonstrates high sensitivity and specificity in identifying multiple ALL sub-types. Furthermore, we introduce cutting-edge telediagnosis software facilitating real-time support for clinicians in promptly and accurately diagnosing various ALL subtypes from microscopic blood smear images. This research aims to enhance leukemia diagnosis efficiency, which is crucial for the timely intervention and managing this life-threatening condition.

The rapid advancement of deep learning technologies has significantly transformed healthcare analytics, particularly in medical data prediction and classification. This study proposes a hybrid Convolutional Neural Network–Long Short-Term Memory (CNN–LSTM) framework for multi-modal healthcare data analysis, integrating medical imaging, structured electronic health records (EHRs), and IoT-generated time-series physiological signals. The proposed architecture combines spatial feature extraction through CNN with temporal dependency modeling via LSTM to enhance predictive accuracy and clinical decision support. A quantitative experimental design was employed, utilizing multi-source healthcare datasets that underwent preprocessing, normalization, and feature engineering prior to model training. The performance of the hybrid model was evaluated using Accuracy, Precision, Recall, F1-Score, AUC-ROC, and Mean Absolute Error (MAE), and compared with conventional machine learning models and standalone deep learning architectures. Experimental results demonstrate that the proposed CNN–LSTM model achieves superior performance, with improved classification accuracy and reduced prediction error, while maintaining strong generalization capability. The findings indicate that integrating spatial and temporal feature learning significantly enhances disease detection, risk stratification, and personalized treatment planning. This approach supports the development of intelligent clinical decision support systems and scalable smart healthcare environments. The proposed framework offers a reliable and efficient solution for advanced healthcare analytics in IoT-enabled systems.

The rapid growth of social media platforms has generated massive volumes of unstructured textual data containing valuable information about public opinions and sentiments. Extracting meaningful insights from this data has become increasingly important for decision-making in various domains, including business, politics, and social analysis. This study aims to evaluate the effectiveness of deep learning techniques for sentiment analysis of social media data, focusing on Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and a hybrid CNN-LSTM model. A quantitative experimental approach is employed, where datasets are preprocessed through text cleaning, tokenization, and feature representation using word embeddings. The models are trained and evaluated using standard performance metrics, including accuracy, precision, recall, and F1-score. The results indicate that all models perform effectively in sentiment classification tasks, with the hybrid CNN-LSTM model achieving the highest performance due to its ability to capture both local textual features and long-term contextual dependencies. This demonstrates that combining CNN and LSTM architectures enhances classification accuracy compared to individual models. Furthermore, the findings confirm that deep learning approaches are more robust in handling the complexity and noisiness of social media data compared to traditional methods. This study contributes to the development of more adaptive and accurate sentiment analysis models and highlights the potential of hybrid deep learning architectures for real-world applications.

In the evolving landscape of agricultural technology, recognizing rice diseases through computational models is a critical challenge, predominantly addressed through Convolutional Neural Networks (CNN). However, the localized feature extraction of CNNs often falls short in complex scenarios, necessitating a shift towards models capable of global contextual understanding. Enter the Vision Transformer (ViT), a paradigm-shifting deep learning model that leverages a self-attention mechanism to transcend the limitations of CNNs by capturing image features in a comprehensive global context. This research embarks on an ambitious journey to refine and adapt the ViT Base(B) transfer learning model for the nuanced task of rice disease recognition. Through meticulous reconfiguration, layer augmentation, and hyperparameter tuning, the study tests the model's prowess across both balanced and imbalanced datasets, revealing its remarkable ability to outperform traditional CNN models, including VGG, MobileNet, and EfficientNet. The proposed ViT model not only achieved superior recall (0.9792), precision (0.9815), specificity (0.9938), f1-score (0.9791), and accuracy (0.9792) on challenging datasets but also established a new benchmark in rice disease recognition, underscoring its potential as a transformative tool in the agricultural domain. This work not only showcases the ViT model's superior performance and stability across diverse tasks and datasets but also illuminates its potential to revolutionize rice disease recognition, setting the stage for future explorations in agricultural AI applications.