cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Rahmadya Trias Handayanto
Contact Email
rahmadya.trias@gmail.com
Phone
-
Journal Mail Official
piksel.unisma@gmail.com
Editorial Address
rogram Studi Teknik Komputer Fakultas Teknik Universitas Islam 45 Jl. Cut Meutia No. 83 Bekasi 17113
Location
Kota bekasi,
Jawa barat
INDONESIA
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic
Published by Universitas Islam 45 Bekasi
ISSN : 23033304     EISSN : 26203553     DOI : https://doi.org/10.33558/piksel
Core Subject : Science,
Computer Science & IT Decision Sciences, Operations Research & Management
Jurnal PIKSEL diterbitkan oleh Universitas Islam 45 Bekasi untuk mewadahi hasil penelitian di bidang komputer dan informatika. Jurnal ini pertama kali diterbitkan pada tahun 2013 dengan masa terbit 2 kali dalam setahun yaitu pada bulan Januari dan September. Mulai tahun 2014, Jurnal PIKSEL mengalami perubahan masa terbit yaitu setiap bulan Maret dan September namun tetap open access tanpa biaya publikasi. p-ISSN: 2303-3304, e-ISSN: 2620-3553. Available Online Since 2018.
Arjuna Subject : Ilmu Komputer (semua kategori) - Ilmu Komputer (Lain-Lain)
Articles 489 Documents
 45   46   47   48   49 
Classifying Half-Unemployment Levels in Indonesian Provinces: A K-Means Approach for Informed Policy Decisions Suhardjono Suhardjono; Hari Sugiarto; Dewi Yuliandari; Adjat Sudradjat; Luthfia Rohimah
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic Vol. 11 No. 2 (2023): September 2023
Publisher : LPPM Universitas Islam 45 Bekasi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33558/piksel.v11i2.7390

Abstract

Half-level unemployment refers to individuals who work part-time and are not fully employed. Increasing the half-poverty rate from year to year can lead to challenges in the lives of these individuals. The issue arising with the rise in the half-poverty rate is the government's difficulty in prioritizing areas that require intervention to address these problems. Consequently, an increase in the half-poverty rate can have adverse consequences. Therefore, it is necessary to categorize underemployment rate data obtained from public sources, specifically from data.go.id, using the widely recognized clustering method known as K-Means. The purpose of this categorization is to identify and classify provinces with a significant prevalence of half-poverty levels. This classification will assist the government in making informed decisions when addressing individuals who meet the half-poverty criteria. The results were obtained by grouping the data from the first to the eighteenth iteration into three categories: 'large' (C1), 'medium' (C2), and 'small' (C3) in terms of half-poverty levels. Group C1 comprises 17 provinces with a high half-poverty rate, while C2 includes only 2 provinces, and C3 covers 16 provinces with a significant half-poverty rate. Based on these findings, it is advisable for the Indonesian government to consider implementing policies aimed at reducing the poverty level by half. Priority should especially be given to the C1 group when creating employment opportunities for the province's residents
Machine Learning-Based Classification for Scholarship Selection Asriyanik Asriyanik; Agung Pambudi
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic Vol. 11 No. 2 (2023): September 2023
Publisher : LPPM Universitas Islam 45 Bekasi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33558/piksel.v11i2.7393

Abstract

University of Muhammadiyah Sukabumi (UMMI) is a university that accepts KIP scholarship every year. However, KIP student applicants always exceed the quota, so it requires a re-selection process to determine KIP Shcolarship Awardee. UMMI does not have a clear method to support decisions in the selection process for KIP Shcholarship Awardee. To solve this problem, a classification modeling process will be carried out from previous data using machine learning algorithms, namely with Decision Tree (DT) and Support Vector Machine (SVM) algorithms. The general method for its development uses the SEMMA method (Sample, Explore, Modify, Model, Assess). Starting with collecting a dataset of KIP recipients studying at UMMI from 2021-2022 which amounted to 519 data with 16 attributes. From the results of exploration, the main attributes that became features for modeling were DTKS Status, P3KE Status, Combined income of father and mother and achievement. These attributes are converted into numeric data for easy data modeling. The results of K-Fold Cross-Validation for the DT model in the case of classification of KIP Kuliah recipients resulted in an accuracy of 78.44% of the entire test dataset, a precision of 0.73107 indicating that 73.11% of the model's predictions were correct, recall (sensitivity level) of 78.45% and an F1 score of 73.20%. The results of modeling and validation with SVM are 80.17% accuracy, 84.44% precision and 80.17% recall. The SVM model shows slightly better in terms of accuracy and precision, both models show competitive performance in classifying KIP scholarship recipients studying at UMMI.
Optimization of Village Budget Plan Selection Based on Priorities Using Method Promethee and Borda Dwi Yanti Laily; Muhammad Dedi Irawan
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic Vol. 11 No. 2 (2023): September 2023
Publisher : LPPM Universitas Islam 45 Bekasi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33558/piksel.v11i2.7890

Abstract

The rapid information technology that is happening now has a significant impact on human life, including various activities that occur at the village office. The village head is responsible for managing funds at the village level including the collection and accountability of funds in accordance with the provisions of law no.33 of 2004 concerning the balance of central and regional governments. This study aims to implement a decision support system by combining two methods, namely the promethee and borda methods in selecting village budget plans in Tegal Sari village based on priority. The method used in this research is the Research and Development (R&D) method. The method used for calculations is the PROMETHEE and Borda methods. The PROMETHEE method is used to manage individual decisions from each decision maker, while the Borda method is used to manage group decisions resulting from the PROMETHEE ranking method. The criteria in this study are Cost, Completion Time, Impact on Regions/Society, Profits from Sustainable Investment, Factors of Interest. The results obtained from this research are Pemb. Health Sarpras as an alternative priority for the village budget.
Enhancing Transformer Performance through Contextual Labeling: A Case Study on Student Mental Health Prediction Mardi Yudhi Putra; Dwi Ismiyana Putri; Rika Apriani; Renaldi Triharsono; Dewi Mufadilah
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic Vol. 14 No. 1 (2026): March 2026
Publisher : LPPM Universitas Islam 45 Bekasi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33558/piksel.v14i1.11800

Abstract

 Early identification of stress and depression among university students is essential to support timely psychological intervention, yet traditional counseling methods often rely on manual, self-initiated reporting that may overlook students experiencing emotional distress. This study aimed to develop a text-based mental-health detection framework using transformer models supported by contextual labeling to analyze student-generated social-media content. The research was conducted through three stages: problem exploration with the Student Affairs Division, data collection from questionnaires and 993 social-media text entries, and comprehensive data preprocessing involving cleaning, normalization, deduplication, and lexicon-based weak labeling. The cleaned dataset was used to fine-tune two transformer architectures—RoBERTa for sequence classification and T5 for text-to-text classification—and to construct a majority-vote ensemble. Model performance was evaluated using accuracy, precision, recall, F1-score, and confusion matrices. The results showed that the T5 model achieved the most balanced performance across all categories, particularly in distinguishing neutral and stress expressions, while RoBERTa and the ensemble exhibited strong prediction bias toward a single class. The findings demonstrated that contextual preprocessing combined with transformer-based modeling effectively supported automated detection of student emotional states. This study concluded that transformer models, especially T5 with contextual labeling, offered a promising foundation for developing early-warning systems that can be integrated into university counseling services and further enhanced through expanded datasets, expert-validated annotations, and explainable-AI components.
Real-Time IoT-Based Monkey Pest Detection Using YOLOv8 for Shallot Farming Deny Rochman Arifatno; Debyo Saptono; M Nur Rois Abid; Susan Siti Nurhaliza
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic Vol. 14 No. 1 (2026): March 2026
Publisher : LPPM Universitas Islam 45 Bekasi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33558/piksel.v14i1.11840

Abstract

Shallot farming is a high-value horticultural sector vital for national food security. Farmers in forest-edge areas face serious losses from monkey pests (Macaca spp.), which can cause rapid, large-scale crop damage. Conventional manual and reactive control methods, such as direct guarding or simple repellents, are often ineffective and unsustainable. This study proposes an IoT-based real-time monkey pest monitoring system using a camera sensor and Raspberry Pi. The system automatically detects monkey presence using the YOLOv8 object detection model and immediately alerts farmers via mobile devices, while activating a buzzer or speaker alarm to deter the animals. The research stages include user needs analysis, system design and implementation with Raspberry Pi 5 as the central processor, field testing, and performance evaluation. The model was trained on approximately 2000 labeled monkey images and achieved 86.3% precision, 85.3% recall, and 90.5% mAP@50. In real-time operation, the system runs at 18–22 frames per second with an overall detection accuracy of 82% and a false positive rate of 8%. The system can distinguish monkeys from humans in the same frame, providing an effective early warning tool for shallot plantations.
Classification of Students Major Preferences at SMKIT Ibnul Qayyim Using a Machine Learning Model Based on Students Knowledge, Skills, and Interests Rajie Al Qadri Anwar; Widyastuti Andriyani
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic Vol. 14 No. 1 (2026): March 2026
Publisher : LPPM Universitas Islam 45 Bekasi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33558/piksel.v14i1.12025

Abstract

This study aims to uncover the dynamics and meanings underlying the major preference process at SMK IT Ibnul Qayyim through the integration of a qualitative approach and the use of machine learning models based on student knowledge, skills, and interests. Major selection is a critical issue in vocational education, as mismatches often occur between student interests and the chosen majors, which can affect learning motivation and job readiness. This study adopts a qualitative case study approach involving ten participants, consisting of guidance and counseling teachers, homeroom teachers, and Grade IX–X students. Data were collected through semi-structured interviews, participatory observation, and document analysis. The data were analyzed using the interactive model of Miles and Huberman, including data reduction, data display, and conclusion drawing. The results reveal three main themes: (1) major determination is still largely influenced by academic achievement rather than skill potential and intrinsic interests; (2) students perceive machine learning-based prediction systems as objective decision-support tools, while emphasizing the importance of teacher involvement in interpreting the results; and (3) the integration of predictive technology with a humanistic approach is more effective in assisting students in determining majors that align with their personal profiles. This analysis aims to evaluate and predict major preferences of vocational high school students in the Software Engineering (Rekayasa Perangkat Lunak/RPL) program based on their academic achievement at the junior secondary school level. The data include scores from core subjects such as Computer Studies, Mathematics, English, Indonesian Language, Arts and Culture, Civic Education, and Social Studies. Two main analytical approaches are employed: Logistic Regression and Random Forest. These methods are selected because each offers distinct strengths in addressing the research objectives, not only in predicting major preferences but also in providing interpretability regarding the factors that influence student decision-making.
Facial Age Estimation on Asian Faces Using SE-ResNeXt50 and Skin Texture Analysis Hanni Deswita; Ben Rahman; Andrianingsih Andrianingsih; Agus Iskandar
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic Vol. 14 No. 1 (2026): March 2026
Publisher : LPPM Universitas Islam 45 Bekasi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33558/piksel.v14i1.12039

Abstract

Image-based facial age estimation is becoming an important component in biometrics and digital dermatology, but many deep learning approaches still rely on global facial features, making them less sensitive to micro changes on the skin surface, particularly on Asian faces which have distinct ageing patterns. This research offers a novel contribution by integrating SE-ResNeXt50 with skin texture analysis to produce more accurate and interpretable age estimations. The dataset used is APPA-REAL, which consists of 7,612 age-labeled Asian face images. After face detection, skin area cropping, size standardisation, and intensity normalisation, visual features were extracted using SE-ResNeXt50, which utilises a channel attention mechanism through Squeeze-and-Excitation blocks to amplify subtle ageing signals. In parallel, this study adds skin texture analysis based on quantitative indicators, namely wrinkle index, tone unevenness, shine proxy, and brightness, to represent the skin microstructure correlated with ageing. The performance of the method is evaluated using Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE). The results show that the combination of an attention-based deep network and skin texture indicators can improve the consistency of age prediction and provide a clearer basis for interpreting changes in skin texture on Asian faces. This finding strengthens the potential for developing an age estimation system that is not only precise but also relevant for digital skin monitoring applications and ageing evaluation.
Face Landmark-Based Drowsiness Warning System for Drivers in Intelligent Transport System to Reduce Accidents with Yolov11 Ndaru Ruseno; Miswanto; Khotimah Nurhaliza Shufiyah
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic Vol. 14 No. 1 (2026): March 2026
Publisher : LPPM Universitas Islam 45 Bekasi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33558/piksel.v14i1.12053

Abstract

The objective of this study was to develop a drowsiness detection system that can help to reduce traffic accidents caused by drivers who are feeling sleepy. The approach employed in this study involved the use of behavioral measures to monitor the driver’s face. The study utilized a prototyping application development model, involving five cycles of testing and refinement. The results showed that the accuracy of the system prototype achieving high accuracy rates 92% compared to previous research with Yolo v11. The system prototype that generates an alert sound for five seconds when a drowsy driver is detected. Testing of the application showed that the drowsiness detection system performed well.
Skill Recommendation System Using User-Based Collaborative Filtering Method Fadli Yandra; Muhammad Iqbal Fadillah; Hari Soetanto
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic Vol. 14 No. 1 (2026): March 2026
Publisher : LPPM Universitas Islam 45 Bekasi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33558/piksel.v14i1.12078

Abstract

The rapidly evolving job landscape in the digital era requires job seekers to continuously adapt to emerging skills driven by technological advancements across various industries. However, many job seekers struggle to keep up with these changing skill demands, while existing job portals often lack features that recommend relevant skills. To address this issue, this study proposes a skill recommendation system based on the User-Based Collaborative Filtering approach, which considers similarities between users’ preferences. Two similarity measurement methods, Log-Likelihood Similarity and Cosine Similarity, are applied and compared to evaluate their effectiveness. The system matches user skill profiles with skill requirements extracted from job vacancy data, where job postings are also treated as user representations. The dataset was collected from the Jobstreet job portal using web scraping techniques, ensuring relevance to real-world job market conditions. The system performance was evaluated using the Hit Rate Matrix. The results show that the Log-Likelihood Similarity method achieved a hit rate of 0.73, outperforming the Cosine Similarity method, which obtained a score of 0.51. This indicates that Log-Likelihood Similarity provides more accurate and relevant skill recommendations. Overall, the proposed system demonstrates the potential to assist job seekers in identifying relevant skills aligned with current market demands, thereby supporting better career decision-making in a competitive and dynamic job environment.
K-Means for IT Asset Segmentation and Demand Forecasting Using Double Exponential Smoothing (DES) Anita aprilianti; Ben Rahman
PIKSEL : Penelitian Ilmu Komputer Sistem Embedded and Logic Vol. 14 No. 1 (2026): March 2026
Publisher : LPPM Universitas Islam 45 Bekasi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33558/piksel.v14i1.12129

Abstract

IT asset management was a crucial aspect in supporting the smooth operation of a company. Poorly planned asset procurement resulted in asset shortages or excesses, which impacted cost efficiency. Frequent problems included the lack of asset grouping based on needs and difficulties in forecasting future IT asset demand. This study aimed to group IT assets using the K-Means method and to forecast IT asset demand using the Double Exponential Smoothing method. Asset grouping was used to assist companies in determining asset procurement priorities. This study used historical IT asset demand data for the period January 2024 to February 2025. The K-Means method was applied to group assets into three categories: submitted, need to consider, and not submitted. The Double Exponential Smoothing method was employed to forecast future IT asset demand by measuring the error rate using Mean Absolute Deviation (MAD), Mean Squared Error (MSE), and Mean Absolute Percentage Error (MAPE). The results showed that the K-Means method helped companies determine IT asset management priorities, while the Double Exponential Smoothing method produced asset demand forecasts with low error rates, thereby supporting more accurate IT asset procurement planning.

Page 48 of 49 | Total Record : 489

 45   46   47   48   49 

Filter by Year

2013 2026


Reset
Filter By Issues
All Issue Vol. 14 No. 1 (2026): March 2026 Vol. 13 No. 2 (2025): September 2025 Vol. 13 No. 1 (2025): Maret 2025 Vol. 12 No. 2 (2024): September 2024 Vol. 12 No. 1 (2024): March 2024 Vol. 11 No. 2 (2023): September 2023 Vol 11 No 2 (2023): September 2023 Vol. 11 No. 1 (2023): March 2023 Vol 11 No 1 (2023): March 2023 Vol 10 No 2 (2022): September 2022 Vol. 10 No. 2 (2022): September 2022 Vol. 10 No. 1 (2022): March 2022 Vol 10 No 1 (2022): March 2022 Vol 9 No 2 (2021): September 2021 Vol. 9 No. 2 (2021): September 2021 Vol 9 No 1 (2021): Maret 2021 Vol. 9 No. 1 (2021): Maret 2021 Vol. 8 No. 2 (2020): September 2020 Vol 8 No 2 (2020): September 2020 Vol 8 No 1 (2020): Maret 2020 Vol. 8 No. 1 (2020): Maret 2020 Vol 7 No 2 (2019): September 2019 Vol. 7 No. 2 (2019): September 2019 Vol. 7 No. 1 (2019): Maret 2019 Vol 7 No 1 (2019): Maret 2019 Vol 6 No 2 (2018): September 2018 Vol. 6 No. 2 (2018): September 2018 Vol 6 No 1 (2018): Maret 2018 Vol. 6 No. 1 (2018): Maret 2018 Vol. 5 No. 2 (2017): September 2017 Vol 5 No 2 (2017): September 2017 Vol 5 No 1 (2017): Maret 2017 Vol. 5 No. 1 (2017): Maret 2017 Vol 4 No 2 (2016): September 2016 Vol. 4 No. 2 (2016): September 2016 Vol. 4 No. 1 (2016): Maret 2016 Vol 4 No 1 (2016): Maret 2016 Vol. 3 No. 2 (2015): September 2015 Vol 3 No 2 (2015): September 2015 Vol. 3 No. 1 (2015): Maret 2015 Vol 3 No 1 (2015): Maret 2015 Vol. 2 No. 2 (2014): September 2014 Vol 2 No 2 (2014): September 2014 Vol. 2 No. 1 (2014): Maret 2014 Vol 2 No 1 (2014): Maret 2014 Vol 1 No 2 (2013): September 2013 Vol. 1 No. 2 (2013): September 2013 Vol 1 No 1 (2013): Januari 2013 Vol. 1 No. 1 (2013): Januari 2013 More Issue