cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Nur Inayah
Contact Email
inprime.journal@uinjkt.ac.id
Phone
+6285280159917
Journal Mail Official
inprime.journal@uinjkt.ac.id
Editorial Address
Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah Jl. Ir H. Juanda No.95, Cemp. Putih, Kec. Ciputat, Kota Tangerang Selatan, Banten 15412
Location
Kota tangerang selatan,
Banten
INDONESIA
InPrime: Indonesian Journal Of Pure And Applied Mathematics
Published by Universitas Islam Negeri Syarif Hidayatullah Jakarta
ISSN : 26865335     EISSN : 27162478     DOI : 10.15408/inprime
Core Subject : Science, Education,
Computer Science & IT Mathematics
InPrime: Indonesian Journal of Pure and Applied Mathematics is a peer-reviewed journal and published on-line two times a year in the areas of mathematics, computer science/informatics, and statistics. The journal stresses mathematics articles devoted to unsolved problems and open questions arising in chemistry, physics, biology, engineering, behavioral science, and all applied sciences. All articles will be reviewed by experts before accepted for publication. Each author is solely responsible for the content of published articles. This scope of the Journal covers, but not limited to the following fields: Applied probability and statistics, Stochastic process, Actuarial, Differential equations with applications, Numerical analysis and computation, Financial mathematics, Mathematical physics, Graph theory, Coding theory, Information theory, Operation research, Machine learning and artificial intelligence.
Arjuna Subject : Matematika - Matematika Terapan
Articles 197 Documents
 10   11   12   13   14 
Machine Learning for Cybersecurity: Web Attack Detection (Brute Force, XSS, SQL Injection) Irfa’issurur, Muhammad; Josaphat, Bony Parulian
InPrime: Indonesian Journal of Pure and Applied Mathematics Vol 7, No 1 (2025)
Publisher : Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15408/inprime.v7i1.41025

Abstract

Security is a top priority in system development, as web portals serve as critical entry points that are frequently targeted by cyber-attacks. Common attack methods include SQL Injection, Cross-Site Scripting (XSS), and Brute Force. The application of machine learning in cybersecurity is growing due to its effectiveness in detecting such threats. This study employs supervised machine learning with six algorithms: K-Nearest Neighbors (KNN), Random Forest, Naïve Bayes, AdaBoost, LightGBM, and XGBoost. The research utilizes the CICIDS2017 and CSE-CICIDS2018 datasets, which contain network traffic data labeled with four categories: Benign, Brute Force, XSS, and SQL Injection. To address the dataset imbalance issue, this study applies Synthetic Minority Oversampling Technique (SMOTE) in conjunction with Principal Component Analysis (PCA) for dimensionality reduction. Performance evaluation is conducted using accuracy, precision, recall, and F1-score metrics, as well as K-Fold Cross Validation, AUC-ROC, and Learning Curve analysis. The results indicate that the Random Forest algorithm achieves the highest classification performance, with an accuracy of 97.77%, precision of 84.07%, recall of 91.96%, and an F1-score of 87.28%. This research contributes by demonstrating the applicability of machine learning in real-time web attack detection, highlighting the advantages of ensemble-based models in handling cybersecurity threats. Additionally, it underscores the importance of dataset preprocessing techniques in enhancing classification performance. Future improvements should focus on optimizing hyperparameters, integrating real-time network traffic analysis, and exploring hybrid models that combine traditional machine learning with deep learning approaches to further enhance detection capabilities.Keywords: Machine learning; Cybersecurity; Web attack detection; Random forest; SMOTE; PCA. Abstrak Keamanan merupakan prioritas utama dalam pengembangan sistem, karena portal web berfungsi sebagai titik masuk penting yang sering menjadi sasaran serangan siber. Metode serangan umum meliputi SQL Injection, Cross-Site Scripting (XSS), dan Brute Force. Penerapan machine learning dalam keamanan siber semakin berkembang karena efektivitasnya dalam mendeteksi ancaman tersebut. Studi ini menggunakan supervised machine learning dengan enam algoritma: K-Nearest Neighbors (KNN), Random Forest, Naïve Bayes, AdaBoost, LightGBM, dan XGBoost. Penelitian ini memanfaatkan kumpulan data CICIDS2017 dan CSE-CICIDS2018, yang berisi data lalu lintas jaringan yang diberi label dengan empat kategori: Benign, Brute Force, XSS, dan SQL Injection. Untuk mengatasi masalah ketidakseimbangan kumpulan data, studi ini menerapkan Synthetic Minority Oversampling Technique (SMOTE) bersama dengan Principal Component Analysis (PCA) untuk pengurangan dimensionalitas. Evaluasi kinerja dilakukan dengan menggunakan metrik akurasi, presisi, recall, dan skor F1, serta K-Fold Cross Validation, AUC-ROC, dan analisis Learning Curve. Hasilnya menunjukkan bahwa algoritma Random Forest mencapai kinerja klasifikasi tertinggi, dengan akurasi 97,77%, presisi 84,07%, recall 91,96%, dan skor F1 87,28%. Penelitian ini berkontribusi dengan menunjukkan penerapan machine learning dalam deteksi serangan web real-time, menyoroti keunggulan model berbasis ensemble dalam menangani ancaman keamanan siber. Selain itu, penelitian ini menggarisbawahi pentingnya teknik praproses dataset dalam meningkatkan kinerja klasifikasi. Peningkatan di masa mendatang harus difokuskan pada pengoptimalan hiperparameter, pengintegrasian analisis lalu lintas jaringan real-time, dan eksplorasi model hybrid yang menggabungkan machine learning tradisional dengan pendekatan deep learning untuk lebih meningkatkan kemampuan deteksi.Kata Kunci: Pembelajaran mesin; Keamanan siber; Deteksi serangan web; Random forest; SMOTE; PCA. 2020MSC: 68T05
Local Metric Dimension of Certain Operation of Generalized Petersen Graph Tadjuddin, Nur Fahri; Nikbakht, Samaneh
InPrime: Indonesian Journal of Pure and Applied Mathematics Vol 7, No 1 (2025)
Publisher : Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15408/inprime.v7i1.41320

Abstract

A subset W of V(G) is called a local resolving set of G if r(u│W)≠r(v│W) for every two adjacent vertices u,v∈V(G). The smallest cardinality of all local resolving sets in G is called the local metric dimension of G, denoted by lmd(G). The local resolving set of G with cardinality lmd⁡(G) is called a local basis of G. In this paper, we present a novel study, a topic that has not been extensively explored in previous research, on the local metric dimension of certain operation of generalized Petersen graph sP_(n,1) and determine the lower and upper bounds of lmd(sP_(n,m)) with n≥3, s≥1, and 1≤m≤⌊(n-1)/2⌋. We also show that the lower bound is sharp.Keywords: Generalized Petersen graph; Local metric dimension; Local resolving set. AbstrakSuatu subset W dari V(G) dikatakan himpunan pembeda lokal dari G jika r(u│W)≠r(v│W) untuk setiap dua titik bertetangga u,v∈V(G). Kardinalitas terkecil dari semua himpunan pembeda lokal di G disebut dimensi metrik lokal dari G, dinotasikan lmd(G). Himpunan pembeda lokal G dengan kardinalitas lmd(G) disebut basis lokal dari G. Pada artikel ini, disajikan sebuah studi baru, topik yang belum dieskplorasi secara ekstensif dalam penelitian sebelumnya, tentang dimensi metrik lokal dari graf hasil operasi tertentu untuk graf Petersen diperumum sP_(n,1) dan menentukan batas atas dan bawah dari lmd(sP_(n,1)) dengan n≥3, s≥1, dan 1≤m≤⌊(n-1)/2⌋. Kami juga menunjukkan bahwa batas bawah tersebut tajam.Kata Kunci: Graf Petersen diperumum, Dimensi metrik local; Himpunan pembeda local. 2020MSC: 05C12, 05C76
Integrating Spatial Autoregressive Exogenous with Ordinary Kriging for Improved Rainfall Prediction in Java: Enhancing Accuracy with Climate Variables and Spatial Autocorrelation Najwa, Sandrina; Pratiwi, Dhanti Aurilia; Ahdian, Muhammad Rhafi; Indriani, Ayu; Mindra, I Gede Nyoman; Falah, Annisa Nur; Ruchjana, Budi Nurani
InPrime: Indonesian Journal of Pure and Applied Mathematics Vol 7, No 1 (2025)
Publisher : Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15408/inprime.v7i1.42070

Abstract

Indonesia is a tropical country with high rainfall influenced by its archipelagic geography and phenomena like El Niño and La Niña. According to the Meteorology, Climatology, and Geophysics Agency (BMKG), La Niña can increase Indonesia's monthly rainfall by 20-40% above normal. Despite numerous existing spatial interpolation methods, there remains a significant research gap in accurately predicting rainfall at unsampled locations, specifically when considering both spatial autocorrelation and multiple climate variables simultaneously. This research proposes Spatial Autoregressive Exogenous Kriging (SAR-X Kriging), a novel hybrid approach that integrates the SAR-X model with Ordinary Kriging to enhance rainfall prediction accuracy. Unlike conventional methods, SAR-X Kriging explicitly captures both spatial dependence and the influence of external climate factors, improving predictive performance. SAR-X Kriging first models spatial dependencies between locations and incorporates exogenous climate variables (surface pressure, air temperature, humidity, wind speed, and solar radiation) to enhance prediction accuracy. It also applies kriging for spatial interpolation. This method was chosen for its robustness in capturing spatial dependence and external influences. The analysis revealed significant spatial dependence across districts/cities in Java Island based on the Moran's Index test. The best SAR-X model, utilizing air temperature and wind speed as exogenous variables, achieved a p-value of 6.0352 × 10-9. Predictions using SAR-X Kriging yielded the lowest Mean Absolute Percentage Error (MAPE) of 3.82%, outperforming the standalone SAR-X method MAPE 4.68% and the Ordinary Kriging method MAPE 3.86%. Practically, these results provide reliable rainfall predictions, enabling better climate-informed decision-making in water resource management, agricultural planning, and flood prevention strategies in Java.Keywords: Climate; Kriging; MAPE; Rainfall; SAR-X. AbstrakIndonesia merupakan negara tropis dengan curah hujan tinggi yang dipengaruhi oleh kondisi geografis kepulauan serta fenomena alam seperti El Niño dan La Niña. Menurut Badan Meteorologi, Klimatologi, dan Geofisika (BMKG), La Niña mampu meningkatkan curah hujan bulanan Indonesia hingga 20-40% di atas normal. Meskipun terdapat berbagai metode interpolasi spasial yang telah dikembangkan, masih terdapat kesenjangan penelitian dalam menghasilkan prediksi curah hujan secara akurat di lokasi yang tidak tersampel, terutama ketika mempertimbangkan secara bersamaan ketergantungan spasial serta pengaruh dari berbagai variabel iklim. Penelitian ini mengusulkan metode bernama Spatial Autoregressive Exogenous Kriging (SAR-X Kriging), sebuah pendekatan hybrid baru yang mengintegrasikan model SAR-X dengan metode Ordinary Kriging untuk meningkatkan akurasi prediksi curah hujan. Tidak seperti metode konvensional, SAR-X Kriging secara eksplisit menangkap ketergantungan spasial serta pengaruh faktor iklim eksternal, sehingga meningkatkan kinerja prediktif. SAR-X Kriging bekerja dengan memodelkan terlebih dahulu ketergantungan spasial antar lokasi, kemudian memasukkan variabel eksogen berupa tekanan permukaan, suhu udara, kelembaban, kecepatan angin, dan radiasi matahari untuk meningkatkan akurasi prediksi, serta terakhir menerapkan teknik kriging untuk interpolasi spasial. Metode ini dipilih karena mampu menangkap secara lebih baik ketergantungan spasial sekaligus pengaruh variabel eksternal dibandingkan metode konvensional. Hasil analisis menunjukkan adanya ketergantungan spasial yang signifikan antar kabupaten/kota di Pulau Jawa berdasarkan uji Moran’s Index. Model SAR-X terbaik diperoleh dengan variabel suhu udara dan kecepatan angin, mencapai nilai p-value sebesar 6.0352 × 10-9. Prediksi menggunakan SAR-X Kriging menghasilkan Mean Absolute Percentage Error (MAPE) sebesar 3,82%, mengungguli metode SAR-X yaitu MAPE 4,68% dan metode Ordinary Kriging yaitu MAPE 3,86%. Secara praktis, hasil ini dapat meningkatkan kualitas prediksi curah hujan yang bermanfaat dalam pengelolaan sumber daya air, perencanaan pertanian, serta strategi mitigasi banjir di Pulau Jawa.Kata Kunci: Iklim, Kriging; MAPE; Curah hujan; SAR-X. 2020MSC: 62H11, 86A32
Completely Closed Filter in BN -Algebra Ramadhan, Andi Rio; Gemawati, Sri; Kartini, Kartini
InPrime: Indonesian Journal of Pure and Applied Mathematics Vol 7, No 1 (2025)
Publisher : Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15408/inprime.v7i1.42170

Abstract

A BN-algebra (A; *,0) is a non-empty set A equipped with a binary operation * and a constant 0, which satisfies the following axioms: (B1)  a*a=0, (B2)  a*0=a, and (BN)  (a*b)*c=(0*c)*(b*a) for all a,b,c ∈A. A subset I of A is called an ideal in A if it satisfies (i) 0∈I, (ii) if b∈I and a*b∈I imply a∈I, for all a,b∈A. This paper presents an original investigation on the completely closed filter in BN-algebra, a topic that has not been extensively explored in previous research. The concepts of filter, closed filter, and completely closed filter in BN-algebra are defined, which can always be associated with the concept of an ideal in BN-algebra. It begins by defining a filter in BN-algebra and then providing additional conditions to make it a closed and completely closed filter. The results show that every filter in BN-algebra has a condition (D), and every non-empty subset of BN1-algebra is a closed filter. Furthermore, every normal ideal in BN-algebra, ideal in Coxeter algebra, and subalgebra in BN1-algebra is a completely closed filter.Keywords: BN-algebra; Completely closed filter; Filter; Ideal. AbstrakBN-Aljabar (A; *,0) adalah himpunan tak kosong A yang dilengkapi dengan operasi biner * dan konstanta 0, yang memenuhi aksioma berikut: (B1) a*a=0,(B2) a*0=a, dan (BN) (a*b)*c=(0*c)*(b*a) untuk setiap a,b,c ∈A. Subhimpunan I dari A disebut ideal di A jika memenuhi: (i) 0∈I, (ii) untuk setiap b∈I dan a*b∈I mengakibatkan a∈I, untuk setiap a,b∈A. Dalam artikel ini, kami menyajikan sebuah studi baru tentang filter tertutup lengkap dalam BN-aljabar, sebuah topik yang belum banyak dieksplorasi dalam penelitian sebelumnya. Konsep filter, filter tertutup, dan filter tertutup lengkap dalam BN-Aljabar didefinisikan, yang mana selalu dapat dikaitkan dengan konsep ideal dalam BN-Aljabar. Dimulai dengan mendefinisikan filter dalam BN-aljabar, kemudian memberikan kondisi tambahan untuk menjadikannya filter tertutup dan filter tertutup lengkap. Hasil yang diperoleh adalah setiap filter dalam BN-Aljabar dengan kondisi (D) dan setiap subset tak kosong dari BN1-aljabar merupakan filter tertutup. Lebih jauh, setiap ideal normal dalam BN-aljabar, ideal dalam Coxeter aljabar, dan subaljabar dalam BN1-aljabar merupakan filter tertutup lengkap.Kata Kunci: BN-aljabar; Filter tertutup lengkap; Filter; Ideal. 2020MSC: 03G25, 03G10
Distance Magic Labeling of Corona Product of Graphs Nadeak, Christyan Tamaro
InPrime: Indonesian Journal of Pure and Applied Mathematics Vol 6, No 1 (2024)
Publisher : Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15408/inprime.v6i1.38317

Abstract

Let G = (V, E) is a graph with order n, and f: V(G) → {1,2,...,n} is a bijection. For any vertex v ϵ V, the sum of f(u) is called the weight of vertex v, denoted by w(v), where N(v)  is the set of neighbors of vertex v. If the labeling f satisfies that there exists a constant k such that w(v)=k, for every vertex v in the graph G, then f is called a distance magic labeling for the graph G. If a graph G has a distance magic labeling, then G is called a distance magic graph. This paper presents a novel result that has not been extensively explored in previous research on the distance magic labeling for the corona product between several families of graphs, such as a complete, cycle, path, and star graph.Keywords: Distance magic labeling; Corona product; Complete graph; Cycle graph; Path graph; Star graph. AbstrakMisalkan G = (V, E) adalah graf berorde n, dan f: V(G) → {1,2,...,n}  merupakan suatu bijeksi. Untuk sebarang titik vϵ V, jumlahan dari f(u) merupakan bobot dari titik v dan dinotasikan dengan w(v), dengan N(v) merupakan himpunan tetangga dari titik v. Jika pelabelan f memenuhi terdapat suatu konstanta k sehingga w(v)=k, untuk setiap titik v yang terdapat pada graf G, maka f disebut sebagai pelabelan ajaib jarak bagi graf G. Jika suatu graf G memiliki pelabelan ajaib jarak, maka G disebut sebagai graf ajaib jarak. Paper ini memberikan hasil yang belum pernah dibahas sebelumnya, yaitu pelabelan ajaib jarak untuk operasi korona antara beberapa keluarga graf, seperti graf lengkap, graf siklus, graf lintasan, dan graf bintang.Kata Kunci: Pelabelan ajaib jarak; Operasi korona; Graf lengkap; Graf siklus; Graf lintasan; Graf bintang. 2020MSC: 
A Multidimensional Approach for Solving Multi-Objective Linear Programming Problems Mohammed, Chiya A.; Ramadan, Ayad M.
InPrime: Indonesian Journal of Pure and Applied Mathematics Vol 7, No 1 (2025)
Publisher : Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15408/inprime.v7i1.45661

Abstract

Solving multi-objective linear programming problems (MOLPP) is a great challenge because it is essential in many real-life problems, especially manufacturing. Choosing the best solution is the goal of the decision-maker to produce a possibility to improve their ability to decide. Multi-dimensional scaling (MDS) gives this capability to the right decision. In this study, we develop the MDS method for (MOLPP) in the work of Mrakhan et al. (2020). The method depends on embedding points in  R^2. Start by building a matrix from a collection of points, and then use clustering to optimize the matrix dimensions and configure the points in R^2. The matrix has (k_1*2) dimensions, where k_1 is the big cluster of the points. Also, a center of points was used to find the scaling points, and then the center of the generated points was used to find a distance from the origin (0, 0). Our proposed algorithm offers a structured, efficient compromise solution for MOLPPs, outperforming traditional scalarization-based methods.Keywords: Comprise solution; Multi-dimensional scaling; Multi-objective linear programming; Optimal advanced; Optimal average; Quadratic average. AbstrakMenyelesaikan masalah pemrograman linier multiobjektif (MOLPP) merupakan tantangan besar karena sangat penting dalam banyak masalah kehidupan nyata, terutama manufaktur. Memilih solusi terbaik adalah tujuan pembuat keputusan untuk menciptakan kemungkinan guna meningkatkan kemampuan mereka dalam mengambil keputusan. Penskalaan multidimensi (MDS) memberikan kemampuan ini untuk keputusan yang tepat. Pada studi ini, akan dikembangkan metode MDS untuk (MOLPP) dalam karya Mrakhan et al. (2020). Metode ini bergantung pada penyematan titik-titik di R^2: dimulai dengan membangun matriks dari kumpulan titik, lalu gunakan pengelompokan untuk mengoptimalkan dimensi matriks dan mengonfigurasi titik-titik di R^2. Matriks memiliki dimensi (k_1*2), dimana k_1 adalah klaster besar titik-titik. Selain itu, titik pusat digunakan untuk menemukan titik penskalaan, kemudian titik pusat tersebut digunakan untuk menemukan jarak dari titik asal (0, 0). Algoritma yang kami usulkan menawarkan solusi kompromi yang terstruktur dan efisien untuk MOLPP, yang mengungguli metode berbasis skalarisasi tradisional.Kata Kunci: Solusi terpadu; Skala multidimensi; Pemrograman linier multiobjektif; Lanjutan optimal; Rata-rata optimal; Rata-rata kuadratik. 2020MSC: 90C29, 90C90.
Fuzzy Unsupervised Artificial Learning Based on Credibilistic Fuzzy C-Means Richard, Kangiama Lwangi; Pierre, Kafunda Katalayi; Blaise, Kabamba Baludikay; Rostin, Mabela Makengo Matendo; Djonive, Munene Asidi
InPrime: Indonesian Journal of Pure and Applied Mathematics Vol 7, No 1 (2025)
Publisher : Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15408/inprime.v7i1.44234

Abstract

This study proposes an unsupervised artificial learning approach based on the Credibilistic Fuzzy C-Means (CFCM) algorithm to enhance the governance and analysis of oil production data. The research focuses on supporting decision-making in managing oil output from the MOTOBA oil field, operated by PERENCO in Moanda, Democratic Republic of Congo, covering the period from 2018 to 2021. The methodology involves structuring and segmenting production data using the CFCM algorithm, which enables the identification of meaningful production patterns despite the presence of uncertainty and imprecision in the data. The analysis identified three distinct clusters: wells with low production, wells with moderate production, and wells with high production. These clusters offer valuable insights into the variability of well performance and provide a basis for optimizing operational strategies. The credibilistic enhancement of traditional fuzzy clustering allows for more effective handling of data uncertainty, resulting in a robust and interpretable model—particularly beneficial in complex and data-limited environments. This clustering framework supports more refined monitoring, resource allocation, and operational planning, making it well-suited for the dynamic nature of oil field management. Furthermore, the methodology demonstrates potential scalability and applicability to other industrial domains facing similar challenges in data quality and decision-making under uncertainty. Ultimately, this work contributes to the advancement of data-driven governance in natural resource management through a rigorous and adaptable analytical approach.Keywords: Artificial learning; Clustering; Credibilist; Fuzzy C-means; Fuzzy logic. AbstrakStudi ini mengusulkan pendekatan pembelajaran buatan tanpa pengawasan berdasarkan algoritma Credibilistic Fuzzy C-Means (CFCM) untuk meningkatkan tata kelola dan analisis data produksi minyak. Penelitian ini berfokus pada dukungan pengambilan keputusan dalam mengelola produksi minyak dari ladang minyak MOTOBA, yang dioperasikan oleh PERENCO di Moanda, Republik Demokratik Kongo, yang mencakup periode 2018 hingga 2021. Metodologi ini melibatkan penataan dan segmentasi data produksi menggunakan algoritma CFCM, yang memungkinkan identifikasi pola produksi yang bermakna meskipun terdapat ketidakpastian dan ketidaktepatan dalam data. Analisis ini mengidentifikasi tiga klaster yang berbeda: sumur dengan produksi rendah, sumur dengan produksi sedang, dan sumur dengan produksi tinggi. Klaster ini menawarkan wawasan berharga tentang variabilitas kinerja sumur dan menyediakan dasar untuk mengoptimalkan strategi operasional. Peningkatan kredibilistik dari pengelompokan fuzzy tradisional memungkinkan penanganan ketidakpastian data yang lebih efektif, menghasilkan model yang kuat dan dapat ditafsirkan—terutama bermanfaat dalam lingkungan yang kompleks dan terbatas data. Kerangka pengelompokan ini mendukung pemantauan, alokasi sumber daya, dan perencanaan operasional yang lebih baik, sehingga sangat sesuai untuk sifat dinamis pengelolaan ladang minyak. Lebih jauh lagi, metodologi ini menunjukkan potensi skalabilitas dan penerapan pada domain industri lain yang menghadapi tantangan serupa dalam kualitas data dan pengambilan keputusan dalam ketidakpastian. Pada akhirnya, karya ini berkontribusi pada kemajuan tata kelola berbasis data dalam pengelolaan sumber daya alam melalui pendekatan analitis yang ketat dan adaptif.Kata Kunci: Pembelajaran buatan; Pengelompokan; Kredibilitas; Fuzzy C-means; Logika Fuzzy. 2020MSC: 68T05, 62H30, 90C90.
A Study on Sentiment Analysis of Public Response to The New Fuel Price Policy In 2022: A Support Vector Machine Approach Putri, Niluh Putu Aprillia Puspitadewi Sudarsana; Angreni, Dwi Shinta; Sudarsana, I Wayan
InPrime: Indonesian Journal of Pure and Applied Mathematics Vol 7, No 1 (2025)
Publisher : Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15408/inprime.v7i1.42717

Abstract

The Indonesian government's decision to raise fuel prices in 2022, following a global surge in crude oil prices, triggered widespread public debate. Understanding public sentiment toward such policy decisions is essential for determining the appropriate timing of implementation while minimizing negative reactions. This study aims to classify public sentiment regarding the fuel price hike using the Support Vector Machine (SVM) algorithm. Data were collected from Twitter through web scraping using the SNScrape library in Python. A total of 3,000 tweets were gathered and underwent preprocessing steps such as case folding, tokenization, stopword removal, and stemming. The classification model was built in Google Colab using the SVM algorithm to categorize tweets as positive (+) or negative (–). Model performance was evaluated using a confusion matrix, achieving an accuracy of 81.0%. The results showed that 63.6% of public responses were negative, while 36.4% were positive. Additionally, it was observed that the accuracy converged to 81.1% as the number of training iterations increased. The findings were presented through word clouds and pie charts to enhance interpretability, and a simple graphical user interface (GUI) was developed for user interaction. The study indicates that the government’s repeated delays in implementing the price adjustment may have reflected sensitivity to public sentiment. This research demonstrates the potential of sentiment classification as a tool for evidence-based policymaking, offering insights into the social dynamics surrounding policy changes. Future research could expand by incorporating multi-class sentiment categories or real-time data for dynamic policy evaluation.Keywords: Fuel price; Public opinion; Sentiment analysis; Social media; SVM. AbstrakKeputusan pemerintah Indonesia untuk menaikkan harga bahan bakar minyak pada tahun 2022 dan disusul oleh lonjakan harga minyak mentah global, memicu perdebatan publik yang meluas. Memahami sentimen publik terhadap keputusan kebijakan tersebut sangat penting untuk menentukan waktu implementasi yang tepat untuk meminimalkan reaksi negatif. Penelitian ini bertujuan untuk mengklasifikasikan sentimen publik terhadap kenaikan harga bahan bakar minyak menggunakan algoritma Support Vector Machine (SVM). Data dikumpulkan dari Twitter melalui web scraping menggunakan pustaka SNScrape dalam bahasa Python. Sebanyak 3.000 tweet dikumpulkan dan dilakukan tahap praproses seperti case folding, tokenization, stopword removal, dan stemming. Model klasifikasi dibangun di Google Colab menggunakan algoritma SVM untuk mengkategorikan tweet sebagai positif (+) atau negatif (–). Kinerja model dievaluasi menggunakan matriks confusion dan mencapai akurasi 81,0%. Hasil penelitian menunjukkan bahwa 63,6% tanggapan publik bersifat negatif, sedangkan 36,4% bersifat positif. Selain itu, akurasi konvergen menjadi 81,1% seiring dengan peningkatan jumlah iterasi pelatihan. Temuan tersebut disajikan melalui word cloud dan diagram pai untuk meningkatkan interpretabilitas, dan graphical user interface (GUI) sederhana dikembangkan untuk interaksi pengguna. Studi ini menunjukkan bahwa penundaan berulang pemerintah dalam menerapkan penyesuaian harga mungkin mencerminkan kepekaan terhadap sentimen publik. Penelitian ini menunjukkan potensi klasifikasi sentimen sebagai alat untuk pembuatan kebijakan berbasis bukti, yang menawarkan wawasan tentang dinamika sosial seputar perubahan kebijakan. Penelitian di masa mendatang dapat diperluas dengan menggabungkan kategori sentimen multikelas atau data waktu nyata untuk evaluasi kebijakan yang dinamis.Kata Kunci: Bahan bakar; Opini public; Analisis sentiment; Mesia social; SVM. 2020MSC: 62H30, 91D30.
Survival Analysis of Stroke Incidence in National Health Insurance Participants from 2015 – 2020 Ilmi, Irfan; Suardi, Lenny
InPrime: Indonesian Journal of Pure and Applied Mathematics Vol 7, No 2 (2025)
Publisher : Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15408/inprime.v7i2.49104

Abstract

This study aims to analyze the survival of stroke patients enrolled in the National Health Insurance (Jaminan Kesehatan Nasional, JKN) program and factors affecting it during the 2015–2020 period. Survival analysis was utilized using the Kaplan-Meier estimator and the Cox Proportional Hazards model. The dataset consisted of 12,773 stroke patients sampled from BPJS Kesehatan administrative records. The results indicate that since being registered as BPJS Kesehatan participants or from the baseline year 2015, stroke patients had an average survival time of 2,264 days, with a 95% confidence interval between 2,240 and 2,287 days. The Cox model revealed that patients aged 18–35, 36–50, 51–65, and >65 had Hazard Ratios (HR) of 1.30, 1.69, 2.47, and 3.52, respectively. Female patients exhibited a lower risk of death (HR = 0.81) than males. Employment segment effects were modest, and regional disparities were observed, with the Eastern region showing a higher risk (HR = 1.29). Comorbidities further increased hazards, with hypertension (HR = 1.70) and diabetes (HR = 2.17) significantly raising mortality risk. As one of the first large-scale survival analyses using JKN national data, this study offers novel evidence on key determinants of stroke outcomes in Indonesia. Its findings highlight critical risk factors and support more targeted, data-driven strategies for stroke prevention under universal health coverage.Keywords: Cox Proportional Hazard; Kaplan-Meier; National Health Insurance Agency; Stroke; Survival Analysis. AbstrakPenelitian ini bertujuan untuk menganalisis survival pasien stroke yang terdaftar dalam program Jaminan Kesehatan Nasional (JKN) dan faktor-faktor yang memengaruhinya selama periode 2015–2020. Metode yang digunakan adalah analisis survival dengan pendekatan Kaplan-Meier dan model Cox Proportional Hazards. Data yang dianalisis diambil dari sampel BPJS Kesehatan peserta JKN selama 2015–2020, yang berjumlah 12.773 pasien. Hasil penelitian menunjukkan bahwa sejak terdaftar sebagai peserta BPJS Kesehatan atau sejak tahun dasar 2015, pasien stroke memiliki waktu survival rata-rata 2.264 hari, dengan interval kepercayaan 95% antara 2.240 dan 2.287 hari. Model Cox mengungkapkan pasien berusia 18–35, 36–50, 51–65, dan >65 memiliki HR masing-masing sebesar 1,30, 1,69, 2,47, dan 3,52. Perempuan memiliki risiko lebih rendah (HR = 0,81) dibandingkan laki-laki. Efek pada segmen pekerjaan relatif kecil, dan disparitas regional teramati, dengan wilayah Timur menunjukkan risiko yang lebih tinggi (HR = 1,29). Komorbiditas semakin meningkatkan risiko, dengan hipertensi (HR = 1,70) dan diabetes (HR = 2,17) secara signifikan meningkatkan risiko mortalitas. Sebagai salah satu analisis survival skala besar pertama yang menggunakan data nasional JKN, studi ini menawarkan bukti baru tentang determinan utama luaran stroke di Indonesia. Temuannya menyoroti faktor risiko kritis dan mendukung strategi pencegahan stroke yang lebih terarah dan berbasis data dalam kerangka jaminan kesehatan semesta.Kata Kunci: Cox Proportional Hazard; Kaplan-Meier; Jaminan Kesehatan Nasional; Stroke; Analisis survival.  2020MSC: 91G05.
Machine Learning for Cybersecurity: Web Attack Detection (Brute Force, XSS, SQL Injection) Irfa’issurur, Muhammad; Josaphat, Bony Parulian
InPrime: Indonesian Journal of Pure and Applied Mathematics Vol. 7 No. 1 (2025)
Publisher : Department of Mathematics, Faculty of Sciences and Technology, UIN Syarif Hidayatullah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15408/inprime.v7i1.41025

Abstract

Security is a top priority in system development, as web portals serve as critical entry points that are frequently targeted by cyber-attacks. Common attack methods include SQL Injection, Cross-Site Scripting (XSS), and Brute Force. The application of machine learning in cybersecurity is growing due to its effectiveness in detecting such threats. This study employs supervised machine learning with six algorithms: K-Nearest Neighbors (KNN), Random Forest, Naïve Bayes, AdaBoost, LightGBM, and XGBoost. The research utilizes the CICIDS2017 and CSE-CICIDS2018 datasets, which contain network traffic data labeled with four categories: Benign, Brute Force, XSS, and SQL Injection. To address the dataset imbalance issue, this study applies Synthetic Minority Oversampling Technique (SMOTE) in conjunction with Principal Component Analysis (PCA) for dimensionality reduction. Performance evaluation is conducted using accuracy, precision, recall, and F1-score metrics, as well as K-Fold Cross Validation, AUC-ROC, and Learning Curve analysis. The results indicate that the Random Forest algorithm achieves the highest classification performance, with an accuracy of 97.77%, precision of 84.07%, recall of 91.96%, and an F1-score of 87.28%. This research contributes by demonstrating the applicability of machine learning in real-time web attack detection, highlighting the advantages of ensemble-based models in handling cybersecurity threats. Additionally, it underscores the importance of dataset preprocessing techniques in enhancing classification performance. Future improvements should focus on optimizing hyperparameters, integrating real-time network traffic analysis, and exploring hybrid models that combine traditional machine learning with deep learning approaches to further enhance detection capabilities.Keywords: machine learning; cybersecurity; web attack detection; random forest; SMOTE; PCA. Abstrak Keamanan merupakan prioritas utama dalam pengembangan sistem, karena portal web berfungsi sebagai titik masuk penting yang sering menjadi sasaran serangan siber. Metode serangan umum meliputi SQL Injection, Cross-Site Scripting (XSS), dan Brute Force. Penerapan machine learning dalam keamanan siber semakin berkembang karena efektivitasnya dalam mendeteksi ancaman tersebut. Studi ini menggunakan supervised machine learning dengan enam algoritma: K-Nearest Neighbors (KNN), Random Forest, Naïve Bayes, AdaBoost, LightGBM, dan XGBoost. Penelitian ini memanfaatkan kumpulan data CICIDS2017 dan CSE-CICIDS2018, yang berisi data lalu lintas jaringan yang diberi label dengan empat kategori: Benign, Brute Force, XSS, dan SQL Injection. Untuk mengatasi masalah ketidakseimbangan kumpulan data, studi ini menerapkan Synthetic Minority Oversampling Technique (SMOTE) bersama dengan Principal Component Analysis (PCA) untuk pengurangan dimensionalitas. Evaluasi kinerja dilakukan dengan menggunakan metrik akurasi, presisi, recall, dan skor F1, serta K-Fold Cross Validation, AUC-ROC, dan analisis Learning Curve. Hasilnya menunjukkan bahwa algoritma Random Forest mencapai kinerja klasifikasi tertinggi, dengan akurasi 97,77%, presisi 84,07%, recall 91,96%, dan skor F1 87,28%. Penelitian ini berkontribusi dengan menunjukkan penerapan machine learning dalam deteksi serangan web real-time, menyoroti keunggulan model berbasis ensemble dalam menangani ancaman keamanan siber. Selain itu, penelitian ini menggarisbawahi pentingnya teknik praproses dataset dalam meningkatkan kinerja klasifikasi. Peningkatan di masa mendatang harus difokuskan pada pengoptimalan hiperparameter, pengintegrasian analisis lalu lintas jaringan real-time, dan eksplorasi model hybrid yang menggabungkan machine learning tradisional dengan pendekatan deep learning untuk lebih meningkatkan kemampuan deteksi.Kata Kunci: pembelajaran mesin; keamanan siber; deteksi serangan web; random forest; SMOTE; PCA. 2020MSC: 68T05

Page 12 of 20 | Total Record : 197

 10   11   12   13   14 

Filter by Year

2019 2025


Reset
Filter By Issues
All Issue Vol 7, No 2 (2025) Vol. 7 No. 2 (2025) Vol. 7 No. 1 (2025) Vol 7, No 1 (2025) Vol 6, No 2 (2024) Vol. 6 No. 2 (2024) Vol. 6 No. 1 (2024) Vol 6, No 1 (2024) Vol 5, No 2 (2023) Vol. 5 No. 2 (2023) Vol 5, No 1 (2023) Vol. 5 No. 1 (2023) Vol. 4 No. 2 (2022) Vol 4, No 2 (2022) Vol 4, No 1 (2022) Vol. 4 No. 1 (2022) Vol. 3 No. 2 (2021) Vol 3, No 2 (2021) Vol. 3 No. 1 (2021) Vol 3, No 1 (2021) Vol 2, No 2 (2020) Vol 2, No 1 (2020) Vol 1, No 2 (2019) Vol 1, No 1 (2019) More Issue