Article Per Year (5 Year)
p-Index From 2021 - 2026
1.897
P-Index
This Author published in this journals
All Journal Populer: Jurnal Penelitian Mahasiswa Jurnal Teknik Mesin, Industri, Elektro dan Informatika Applied Information Technology and Computer Science (AICOMS)
Ghufron Ghufron
Universitas Islam Sultan Agung
Humanities Computer Science & IT Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Education Electrical & Electronics Engineering Industrial & Manufacturing Engineering Languange, Linguistic, Communication & Media Social Sciences Other

Published : 14 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 14 Documents
Search

 1   2 
Prediksi Harga Saham Sektor Energi di BEI Menggunakan Model Multimodal LSTM dengan Integrasi Fitur Numerik dan Sentimen Berita Pasar Modal di Indonesia Naufal Lathifan Yumna; Ghufron Ghufron
Applied Information Technology and Computer Science (AICOMS) Vol 5 No 1 (2026): AICOMS
Publisher : Pengelola Jurnal Politeknik Negeri Ketapang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58466/jm8dtf93

Abstract

The movement of energy sector stock prices exhibits high volatility, which is influenced by historical data and news sentiment. This research evaluates the performance of stock price prediction by integrating numerical data and economic news texts using a multimodal Long Short-Term Memory (LSTM) architecture. The data utilized includes the stock prices of ADRO, PGAS, and INDY from the 2021 to 2026 period, alongside Indonesian-language economic news. Sentiment extraction from the news texts was conducted using the IndoBERT model. The results indicate that the IndoBERT model achieved an accuracy and F1-score of 83%. The evaluation of the unimodal model (historical data only) yielded a Mean Absolute Percentage Error (MAPE) of 3.62% for ADRO, 3.17% for PGAS, and 5.90% for INDY. Meanwhile, the multimodal model, which combined numerical and sentiment features, resulted in a MAPE of 4.00% (ADRO), 5.46% (PGAS), and 8.47% (INDY). In conclusion, the unimodal LSTM model proved to be effective; however, the integration of sentiment features in the multimodal scheme did not provide a significant improvement in accuracy due to the highly volatile nature of the stocks.
Perbandingan Metode Temporal Fusion Transformer (TFT) dan Long Short-Term Memory (LSTM) Dalam Prediksi Harga Saham Indonesia Berbasis Data Teknikal dan Fundamental Nurhasan Nurhasan; Ghufron Ghufron
Applied Information Technology and Computer Science (AICOMS) Vol 5 No 1 (2026): AICOMS
Publisher : Pengelola Jurnal Politeknik Negeri Ketapang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58466/jmhe8c20

Abstract

The Indonesian capital market exhibits a high level of volatility, requiring stock price prediction models that are accurate and adaptive. Conventional forecasting models have limitations in capturing nonlinear patterns and multivariate relationships within stock time series data. This study compares the performance of Long Short-Term Memory (LSTM) and Temporal Fusion Transformer (TFT) models in predicting stock prices of LQ45 index companies, namely BBRI, TLKM, and ADRO, with a 7-day forecasting horizon. Both models were trained using 12 combined technical and fundamental features with a data split ratio of 70:15:15. Model evaluation was conducted using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE). The results indicate that the TFT model achieved better performance on BBRI with a MAPE of 0.74% and TLKM with a MAPE of 0.91%, while also demonstrating faster training convergence compared to LSTM. In contrast, the LSTM model outperformed TFT on ADRO with a MAPE of 2.71%, which exhibited a relatively consistent trend pattern. Overall, TFT proved to be more effective for stocks with complex multivariate dynamics, whereas LSTM remained competitive for stocks with more stable trend patterns. The selection of prediction models should therefore consider the volatility characteristics and movement patterns of each stock issuer
Analisis Sentimen menggunakan IndoBERT dan Tren Topik Keluhan Pasien pada Ulasan Google Maps Rumah Sakit Menggunakan Latent Dirichlet Allocation Naufal Muhammad Afif; Ghufron Ghufron
Applied Information Technology and Computer Science (AICOMS) Vol 5 No 1 (2026): AICOMS
Publisher : Pengelola Jurnal Politeknik Negeri Ketapang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58466/m86x0x75

Abstract

Patient satisfaction is a crucial indicator of hospital quality, yet management often focuses solely on star ratings that fail to explain the root causes of issues. This study develops a hybrid Natural Language Processing (NLP) model using IndoBERT for sentiment classification of Google Maps reviews. Reviews classified as negative sentiment are then filtered and processed using the Latent Dirichlet Allocation (LDA) method to uncover hidden themes within patient complaints. The test results show that the IndoBERT model achieves exceptionally high performance, with an accuracy of 95.23%, precision of 95.22%, recall of 95.23%, and an F1-score of 95.22%. The LDA analysis successfully identifies 10 optimal topics, which are categorized into five main complaint categories: time efficiency, medical services, facilities/parking, administrative procedures, and specialist services. The integration of IndoBERT and LDA proves effective in transforming raw digital reviews into strategic information for the automated evaluation of hospital service quality.
Prediksi Kelulusan Tepat Waktu Program Pendidikan Profesi Dokter Gigi Menggunakan Algoritma Random Forest Ghufron Ghufron; Dedie Nugroho; Andi Riansyah; Dedy Kurniadi
Jurnal Teknik Mesin, Industri, Elektro dan Informatika Vol. 5 No. 1 (2026): Jurnal Teknik Mesin, Industri, Elektro dan Informatika
Publisher : Lembaga Pengembangan Kinerja Dosen

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55606/jtmei.v5i1.6112

Abstract

The number of dentists in Indonesia as of May 2025 remains relatively low compared to the population size. The government expects higher education institutions as producers of dental graduates to increase the number of graduates in order to meet the national healthcare workforce demand. This condition has become an important concern for Dental Professional Education Programs, particularly the Faculty of Dentistry at Sultan Agung Islamic University, not only in supporting accreditation achievement but also in contributing to the increase in the number of dentists in Indonesia. Various efforts have been implemented, such as clinical supervision, evaluation of clinical learning outcomes, and comprehensive examinations. However, in the implementation of the Dental Profession Student Competency Examination (UKMP2DG), there are still students who fail to pass the examination. This study offers an innovative solution by utilizing Machine Learning technology to predict student graduation based on academic data using the Random Forest algorithm. The research data include UKMP2DG participants from period 1 of 2020 to period 1 of 2025. The results show that the Random Forest model provides good predictive performance with an accuracy of 83.19%, precision of 88.51%, recall of 89.53%, and F1-score of 89.02%. In addition, the most influential feature in predicting graduation was the undergraduate Grade Point Average (GPA), with a contribution level of 41.57%. The findings of this study are expected to serve as a basis for academic decision-making and assist institutions in improving student development strategies to enhance the UKMP2DG graduation rate.
Sistem Deteksi Dini Penyakit Stroke Menggunakan Algoritma Support Vector Machine (SVM) Ahmad Labib Zainuddin; Ghufron Ghufron
Populer: Jurnal Penelitian Mahasiswa Vol. 5 No. 3 (2026): September : Populer: Jurnal Penelitian Mahasiswa
Publisher : Universitas Maritim AMNI Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58192/populer.v5i3.5452

Abstract

Stroke is one of the most dangerous non-communicable diseases and a leading cause of death and long-term disability worldwide, including in Indonesia. Early detection of stroke risk is essential to prevent more severe conditions and to accelerate medical treatment. This study aims to develop an early detection system for stroke using the Support Vector Machine (SVM) algorithm. The dataset was obtained from Kaggle (Stroke Prediction Dataset), consisting of 5,110 patient records with features such as age, gender, hypertension, heart disease, work status, body mass index (BMI), average glucose level, and smoking status. The data underwent preprocessing, including data cleaning, categorical encoding, normalization using a Standard Scaler, and an 80:20 train-test split. The SVM model was tested using four kernel types (linear, polynomial, RBF, and sigmoid) to obtain the most optimal configuration. The RBF kernel achieved the best performance, with 91% accuracy during kernel comparison and a final accuracy of 84.58%, average precision of 0.85, average recall of 0.85, average F1-score of 0.85, and an AUC of 0.92. The system was implemented as a Streamlit-based dashboard so that it can be used as an early screening tool for stroke risk by medical personnel and the general public.
Robot Pelayan Restoran Berbasis Nav2 dan YOLOv11 dengan Sistem Navigasi Cerdas serta Deteksi Objek Real-Time Muhammad Ridwan; Ghufron Ghufron
Populer: Jurnal Penelitian Mahasiswa Vol. 5 No. 3 (2026): September : Populer: Jurnal Penelitian Mahasiswa
Publisher : Universitas Maritim AMNI Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58192/populer.v5i3.5462

Abstract

The context-blind limitation of 2D LiDAR distance sensors in distinguishing different types of obstacles poses a major challenge in the navigation of restaurant service mobile robots. This study aims to design and implement an intelligent autonomous navigation system based on the Navigation 2 (Nav2) framework in ROS 2, integrated with real-time visual perception using YOLOv11-based Deep Learning. The system is built using a dual-processor architecture, in which an NVIDIA Jetson Orin Nano acts as the High-Level Controller, while an ESP32-S3 microcontroller serves as the Low-Level Controller. The YOLOv11 module is trained to classify human postures into person_standing and person_sitting, which are used as additional semantic information in the Nav2 decision-making process managed by a custom Behavior Tree. Experimental results show that both systems achieved a Success Rate of 100% with zero collision incidents. The Intelligent Navigation system achieved an average Task Completion Time of 12.83 seconds, resulting in an average reduction of 1.70 seconds compared to the Standard Navigation system.
Implementasi Prediksi Stunting Menggunakan Algoritma Xgboost dan Smote Berdasarkan Data BKKBN Sabila Wahyu Pratama; Ghufron Ghufron
Populer: Jurnal Penelitian Mahasiswa Vol. 5 No. 3 (2026): September : Populer: Jurnal Penelitian Mahasiswa
Publisher : Universitas Maritim AMNI Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58192/populer.v5i3.5468

Abstract

The management and intervention of stunting cases in toddlers still face major challenges due to the high class imbalance in risk assessment data, where the population of normal toddlers significantly dominates the at-risk class. This condition often causes machine learning classification models to suffer from prediction bias and reduced accuracy. This study aims to build a Machine Learning classification model based on the XGBoost algorithm combined with the Synthetic Minority Over-sampling Technique (SMOTE) to predict stunting risk and analyze its primary key determinants (Feature Importance) using Stunting Risk Family (KRS) data from BKKBN Central Java. The dataset comprises 7,605 family data records with evaluation performed on 2,771 pure test data points. The evaluation results demonstrate that the combination of XGBoost and SMOTE achieved an Accuracy of 88.20% and a Precision of 99.91%. Furthermore, Gain analysis confirms that Gender (score 0.3232) and Sanitation/Latrine Facilities (score 0.2182) are the two most dominant risk factors for stunting. These findings demonstrate the potential of leveraging artificial intelligence systems to support efficient decision-making for healthcare workers within a smart governance framework.
Identifikasi Pelanggaran Lalu Lintas Kendaraan Mobil terhadap Larangan Putar Balik menggunakan Algoritma You Only Look Once (YOLOv11) Tiza Farhani Maulia; Dedy Kurniadi; Ghufron Ghufron
Populer: Jurnal Penelitian Mahasiswa Vol. 5 No. 3 (2026): September : Populer: Jurnal Penelitian Mahasiswa
Publisher : Universitas Maritim AMNI Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58192/populer.v5i3.5487

Abstract

The purpose of this study is to identify and analyze traffic violations involving cars that violate U-Turn prohibitions using the You Only Look Once (YOLOv11) algorithm. Illegal U-Turns in restricted areas are a primary cause of traffic congestion and accidents. This study develops an automated, real-time U-Turn violation monitoring system by integrating the YOLOv11n model, ByteTrack tracking algorithm, and Streamlit interface using 2-zone polygon logic. The dataset covers 7 vehicle classes evaluated on 1,072 validation images. Experimental results show that the YOLOv11n model achieves a mean Average Precision (mAP@50) of 66.5%, Precision of 68.4%, and Recall of 64.2%. With an inference speed of 3.2 ms per frame (169 FPS), the 2-zone logic integrated into Streamlit accurately detects illegal maneuvers, updates real-time metrics, and minimizes false alarms without frame lag. The proposed system demonstrates the potential of combining object detection, vehicle tracking, and zone-based logic for automated monitoring of U-Turn violations in restricted areas.
Penerapan Extratrees Classifier dan Light Gradient Boosting Machine untuk Rekomendasi Perdagangan Kripto Berdasarkan Filter Volatilitas Pasar Aditya Bachtiar; Ghufron Ghufron
Populer: Jurnal Penelitian Mahasiswa Vol. 5 No. 3 (2026): September : Populer: Jurnal Penelitian Mahasiswa
Publisher : Universitas Maritim AMNI Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58192/populer.v5i3.5503

Abstract

The cryptocurrency market is characterized by fluctuating, non-stationary prices and rapidly changing volatility, which makes market-condition analysis complex. This study applies a hybrid machine learning approach combining ExtraTrees Classifier and LightGBM Regressor to support Ethereum trading analysis. Daily ETH/USDT data were collected from the Binance Public API for 1 January 2021 to 31 December 2025, totaling 1,826 raw observations. After preprocessing, feature engineering, and target formation, 1,622 observations were used and split chronologically into 70% training and 30% testing data. ExtraTrees classifies the next-period market condition into Low Volatility and High Volatility, while LightGBM estimates the next-period price change as a price delta. ExtraTrees achieved 89.12% test accuracy, exceeding the 65.91% majority baseline. On the training data, LightGBM obtained MAE of US$47.41, RMSE of US$75.68, R² of 0.9930, and Directional Accuracy of 93.04%. The outputs of both models were integrated using fusion logic with 75% volatility score and 25% magnitude score, producing Prepare to Trade, No Trade Zone (Sideways), and High Vol (No Trade) recommendations. The three-state accuracy reached 73.31%, while binary volatility accuracy reached 90.55%. SHAP was used to explain feature contributions, and the final model was implemented in a Streamlit-based Decision Support System.
Deteksi Real-Time dan Pemetaan Lokasi Jalan Berlubang menggunakan Algoritma YOLOv11 Trianando Mohammad Dahlan Wirayudha; Dedy Kurniadi; Ghufron Ghufron; Muhamad Miftahul Huda
Populer: Jurnal Penelitian Mahasiswa Vol. 5 No. 3 (2026): September : Populer: Jurnal Penelitian Mahasiswa
Publisher : Universitas Maritim AMNI Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58192/populer.v5i3.5513

Abstract

Road damage, particularly potholes, in Genuk District, Semarang City, is dynamic due to tidal flooding and heavy logistics vehicle loads. Conventional monitoring based on manual surveys and citizen reports is relatively slow and may delay repair responses. This study develops a real-time pothole detection and mapping system by integrating the YOLOv11 algorithm with a Web-GIS interface. The system uses a distributed architecture in which a Python-based local inference layer processes video streams and synchronizes detected potholes with GPS coordinates. Image evidence is uploaded to Cloudinary, while spatial coordinates are stored in a PostGIS-configured PostgreSQL database. The resulting spatial data are visualized through an interactive map developed using TypeScript and Leaflet.js on Vercel serverless infrastructure. The YOLOv11 model was trained for 300 epochs using an NVIDIA Tesla T4 GPU. Evaluation achieved 97.4% precision, 90.4% recall, and 95.1% mAP@50, with an inference latency of 9.3 milliseconds per frame, equivalent to approximately 107 FPS. Functional testing confirmed reliable data transmission and automatic deduplication using the ST_DWithin spatial query. The system provides accurate, responsive, and reliable support for mapping potholes and accelerating road infrastructure repair processes.
Co-Authors Aditya Bachtiar Ahmad Labib Zainuddin Dedie Nugroho Dedy Kurniadi Hadana Maulana Ibra Andnan Faiz Sempana Imam Much Ibnu Subroto Mahmudi Husain Hasbullah Muhamad Miftahul Huda Muhammad Ridwan Muhammadun Hanif Mustafa, Mustafa Naufal Lathifan Yumna Naufal Muhammad Afif Nurhasan Nurhasan Poetro, Bagus Satrio Waluyo Riansyah, Andi Sabila Wahyu Pratama Tiza Farhani Maulia Trianando Mohammad Dahlan Wirayudha