Claim Missing Document
Check
Articles

Found 28 Documents
Search

Rice Quality Identification Built on Indonesian Food Standards Based on Electronic Nose using Naïve Bayes Algorithm Muhammad Jauhar Vikri; Ifnu Wisma Dwi Prastya; Ucta Pradema Sanjaya; Mula Agung Barata
Journal of Innovation and Technology Polbeng Series on Informatics (INOVTEK Polbeng - Seri Informatika) Vol. 10 No. 1 (2025): March
Publisher : P3M Politeknik Negeri Bengkalis

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35314/0y0xct32

Abstract

Rice is a staple food in Indonesia, where its quality is regulated by the National Food Standards outlined in National Food Agency Regulation No. 2 of 2023 on Rice Quality and Labeling Requirements. Rice is classified into four grades: premium, medium 1, medium 2, and medium 3. The widespread practice of mislabeling lower-quality rice as a premium through repackaging highlights the critical need for quality control measures. An electronic nose (e-nose) is a reliable device for food quality control. Previous studies have demonstrated its ability to classify rice into two quality grades with 80% accuracy. This study uses exponential data transformation and the Naive Bayes algorithm to enhance the classification accuracy for four rice quality grades according to national standards. The methodology includes signal acquisition, feature extraction using statistical parameters, exponential data transformation, classification, and performance evaluation. The results show that exponential data transformation improves classification accuracy to 97%. This technology can be implemented for automated quality control in milling facilities, storage warehouses, and distribution centres, ensuring consistent rice quality while enhancing supply chain efficiency. The e-nose-based model offers a fast and reliable solution, minimising reliance on human operators.
Pelatihan Pembuatan Lilin Abadi sebagai Upaya Peningkatan Keterampilan dan Kemandirian Ekonomi Masyarakat Ifnu Wisma Dwi Prastya
Nawasena Bhakti Vol. 2 No. 1 (2026): Nawasena Bhakti: Jurnal Pengabdian Masyarakat
Publisher : Badan Usaha Milik Desa Berkaho Pungpungan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.64084/nawasenabhakti.v2i1.179

Abstract

The economic problems of rural communities and the low utilization of household waste into economically valuable products remain challenges in efforts to improve community welfare. Used cooking oil waste, which is generally discarded directly, has the potential to cause environmental pollution if not managed properly. This community service activity aims to enhance the skills and economic independence of the community through training in making eternal candles in Gunungrejo Village, Kedungpring District, Lamongan Regency. The method of implementing the activities was carried out through several stages, namely preparation, socialization, practical training, evaluation, and follow-up community assistance. The training was conducted on December 24, 2024, using a hands-on method for processing used cooking oil into eternal candles that have utility and market value. The results of the activities showed an increase in the knowledge and skills of the community in utilizing household waste into creative products, as well as the emergence of motivation to develop home-based businesses centered on eternal candles. This activity has a positive impact on increasing environmental awareness, productive skills, and opportunities for sustainable economic independence of the community.
TikTok Sentiment Analysis on Koperasi Merah Putih Using SVM and ANN Meliysa Pasa Bagna Aprilia Said; Kholifatus Sholihah; Ifnu Wisma Dwi Prastya; Afril Efan Pajri
Journal of Applied Informatics and Computing Vol. 10 No. 4 (2026): August 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i4.13394

Abstract

TikTok has become a relevant social media source for observing public responses to public issues, including the Koperasi Merah Putih program. This study compares Support Vector Machine (SVM) and Artificial Neural Network (ANN) for classifying sentiment in TikTok comments. The dataset was obtained through TikTok comment scraping and consisted of 25,669 raw comments. After removing empty comments and applying preprocessing stages consisting of cleaning, case folding, normalization, tokenization, stopword removal, and stemming, 21,026 comments were used for sentiment analysis. Sentiment labels were generated automatically using a lexicon-based sentiment labeling approach and grouped into three classes: positive, negative, and neutral. TF-IDF was used for feature extraction with a maximum of 5,000 features and unigram-bigram representation. The dataset was split into training and testing sets with an 80:20 ratio, while Stratified K-Fold Cross Validation and SMOTE were applied to strengthen evaluation and address class imbalance. The results show that SVM achieved the best overall performance before SMOTE with an accuracy of 86.66% and an F1-score of 86.76%. ANN achieved an accuracy of 85.31% before SMOTE and improved slightly after SMOTE to 85.47%. These findings indicate that SVM is more stable for TF-IDF-based TikTok comment classification, while SMOTE can improve ANN performance slightly but does not always increase all models equally.
Penguatan Literasi Kecerdasan Buatan (Artificial Intelligence) bagi Mahasiswa untuk Meningkatkan Adaptasi di Era Digital Ifnu Wisma Dwi Prastya; Lisa Aminatul Mukarromah; Mula Agung Barata
Karsa Bhakti : Journal of Community Service Vol 2 No 2 (2026): Karsa Bhakti : Journal of Community Service
Publisher : CV. Agrapana Media

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65789/kb-jocs.v2i2.106

Abstract

The development of Artificial Intelligence (AI) technology in the digital era has changed students' learning patterns, but its utilization has not been fully accompanied by adequate conceptual literacy and ethical understanding. Students generally have used various AI-based platforms, but this is still limited to practical use without understanding the working principles, information validity, and academic limitations. This community service activity aims to improve students' artificial intelligence literacy as an effort to adapt to the demands of learning and the working world in the digital era. The implementation method of the activity includes the preparation stage, socialization, practice-based training using Giri AI, and evaluation through pretest and posttest. The activity was held on February 23, 2026, at the Faculty of Sharia and Adab Building of Nahdlatul Ulama Sunan Giri University, involving 30 students from UNUGIRI, IKIP, UNIGORO, and Universiti Malaysia Pahang Al-Sultan Abdullah. The implementation results showed a significant increase in participants' AI literacy, demonstrated by the average score rising from 58.8 in the pretest to 86.0 in the posttest, an increase of approximately 46%. This training not only enhanced the skills in using AI academically but also built ethical awareness and critical thinking abilities among students in responsibly leveraging technology in the era of digital transformation.
Klasifikasi Jejak Bau Rokok pada Napas Berdasarkan Interval Waktu Pasca-Merokok Menggunakan Electronic Nose dan Pembelajaran Mesin Ifnu Wisma Dwi Prastya; Mula Agung Barata; Dwi Issadari Hastuti; Rahmat Tegar Patriot Hari Lambang; Muhammad Ahsanul Fikri
JURIKOM (Jurnal Riset Komputer) Vol. 13 No. 5 (2026): Oktober 2026
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/jurikom.v15i5.10146

Abstract

Paparan asap rokok di rumah tangga merupakan faktor risiko bronkopneumonia pada balita, sementara anggapan bahwa napas perokok aman setelah jeda singkat belum diuji pada menit-menit awal pasca-merokok dengan perangkat berbiaya rendah. Penelitian ini menguji electronic nose (E-Nose) berbasis empat sensor gas semikonduktor seri MQ untuk mengklasifikasikan jejak bau rokok pada napas ke dalam empat kelas: perokok 0, 1, dan 10 menit setelah menghabiskan satu batang rokok filter, serta non-perokok. Data berasal dari 12 responden (6 perokok, 6 non-perokok) dengan satu sesi perekaman per responden. Pra-pemrosesan terbaik (Z-Score dan Wavelet Transform) dan ekstraksi ciri statistik per jendela menghasilkan 112 sampel dengan 24 fitur. Random Forest, K-Nearest Neighbor, XGBoost, dan stack deep neural network dievaluasi dengan validasi antar-sesi leave-one-session-out sebagai ukuran utama dan Stratified K-Fold (k=5) tingkat jendela sebagai pembanding. Pada validasi antar-sesi, akurasi tertinggi 74,11% dicapai K-Nearest Neighbor, sedangkan Random Forest mencapai 59,82%, jauh di bawah akurasinya pada Stratified K-Fold (94,64%) yang terbukti terlalu optimistis karena jendela dari sesi yang sama masuk ke data latih dan data uji. Random Forest tetap memisahkan napas perokok dan non-perokok tanpa kesalahan pada kedua skema, tetapi pembedaan lama jeda waktu belum dapat digeneralisasi ke responden baru. Kontribusi utama penelitian ini adalah bukti awal kelayakan E-Nose berbiaya rendah untuk penapisan keberadaan jejak rokok pada napas serta penegasan pentingnya validasi berbasis responden. Dengan dua responden per interval, temuan ini masih bersifat pembuktian konsep.
Comparison of Feature Selection Methods in Classifying Poverty Levels in Indonesia Using Comparative Machine Learning Methods Kamayanti, Nuniska Dwi; Dwi Prastya, Ifnu Wisma; Sahri, Sahri
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12604

Abstract

Poverty classification requires models capable of handling multidimensional data and imbalanced class distributions. This study aims to develop and compare several machine learning algorithms for classifying poverty levels in Indonesia, as well as to analyze the impact of feature selection and reduction methods on model performance. The study employs a comparative approach using a secondary dataset consisting of 514 districts/cities with socio-economic indicators and a binary target variable. The methodology includes data preprocessing, the application of Chi-Square, Pearson Correlation, and Principal Component Analysis (PCA), and the handling of imbalanced data using the Synthetic Minority Oversampling Technique (SMOTE). Modelling is conducted using Random Forest, Support Vector Machine (SVM), Logistic Regression, and Artificial Neural Network (ANN), with evaluation performed using Stratified K-Fold Cross Validation and metrics including accuracy, precision, recall, and F1-score. The results indicate that Chi-Square and Pearson Correlation outperform PCA, with Random Forest achieving the best performance, attaining an accuracy of 0.9854 and an F1-score of 0.9507, while effectively detecting the minority class. Therefore, the combination of Chi-Square and Random Forest is identified as the most effective approach in this study, as it produces a model that is accurate, stable, and capable of handling imbalanced data.
Attention-Enhanced Multivariate Forecasting for Intelligent Microservice Autoscaling Saifuddin, Nur; Barata, Mula Agung; Dwi Prastya, Ifnu Wisma
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12662

Abstract

Proactive autoscaling in cloud-native microservices requires anticipatory decisions because reactive controllers often lag under abrupt workload shifts. This study aims to improve autoscaling decision quality through a two-stage machine learning pipeline. The research adopts an experimental design using production-grade microservice traces, with strict time-respecting train-validation-test splits and training-only fitting for preprocessing and oracle-threshold estimation to prevent leakage. In the first stage, multivariate forecasting models predict future CPU and memory utilization from engineered temporal features. In the second stage, the predicted signals are combined with observed features to classify three autoscaling actions: scale down, hold, and scale up. Benchmarking shows recurrent neural models are strong baselines, while an attention-enhanced encoder-decoder performs best. The best Bahdanau-attention model with residual connection reduces test CPU RMSE from 0.030977 to 0.028924 and memory RMSE from 0.010322 to 0.005452 relative to the strongest BiLSTM baseline. For decision learning, the optimized Extreme Gradient Boosting model using prediction-augmented features achieves an accuracy of 0.950602 and an F1 score of 0.951026. Supporting downstream validation also yields lower SLO violation rates than horizontal and vertical baselines while maintaining zero downtime in the evaluated scenarios. These findings indicate that improving forecasting quality and explicitly transferring predictive signals to the decision stage strengthens proactive autoscaling performance.
Impact of Normalization Methods on Credit Risk Classification with Imbalanced Data Rizky Ramadhan, Ahmad Abil; Nur Afifa, Gusti Wiwik; Barata, Mula Agung; Dwi Prastya, Ifnu Wisma; Pajri, Afril Efan
Journal of Applied Informatics and Computing Vol. 10 No. 5 (2026): October 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i5.13365

Abstract

Credit risk classification is an essential task in the financial sector to minimize loan default losses and maintain financial system stability. One of the main challenges in building accurate credit risk models is class imbalance, where the number of non-default cases significantly outnumbers default cases, causing classifiers to be biased toward the majority class. This study investigates the effect of two normalization methods, Z-Score and Min-Max, on the performance of Support Vector Machine (SVM) and Artificial Neural Network (ANN) classifiers for imbalanced credit risk prediction. Six experimental scenarios were designed, incorporating the Synthetic Minority Over-sampling Technique (SMOTE), Chi-Square feature selection, and GridSearchCV hyperparameter tuning. Experiments were conducted on a dataset of 4,801 loan records with a class imbalance ratio of 77:23, divided into 80% training and 20% testing using stratified sampling. Evaluation metrics include Accuracy, Precision, Recall, F1-Score, AUC-ROC, Matthews Correlation Coefficient (MCC), and Balanced Accuracy, with F1-Score as the primary metric given the imbalanced class distribution. The results demonstrate that Z-Score normalization, when applied in isolation, consistently outperforms Min-Max across both classifiers, with ANN achieving an accuracy of 95.42% and F1-Score of 0.9332 on the Z-Score scenario without SMOTE. However, the fully optimized pipeline combining Min-Max normalization, Chi-Square feature selection, SMOTE, and hyperparameter tuning surpassed this result, yielding the best overall performance with an ANN accuracy of 95.73%, F1-Score of 0.9393, and AUC-ROC of 0.9610. This indicates that Min-Max’s sensitivity to outliers can be effectively mitigated through appropriate feature selection and hyperparameter tuning. These findings provide empirical guidance for selecting optimal preprocessing strategies in imbalanced financial classification tasks.