IRPI Publisher Journals (Institute of Research and Publication Indonesia)
Not a member yet
1012 research outputs found
Sort by
Penerapan Machine Learning untuk Prediksi Kenaikan Harga Beras Premium Menggunakan Algoritma Regresi Linier: Application of Machine Learning for Premium Rice Price Increase Prediction Using Linear Regression Algorithm
No full textKetidakstabilan harga beras premium sebagai komoditas pangan pokok memerlukan solusi prediksi yang akurat untuk membantu perencanaan ekonomi. Penelitian ini menerapkan algoritma Machine Learning, yaitu Regresi Linier, untuk memprediksi kenaikan harga beras premium. Model dilatih menggunakan data historis harga dan dievaluasi kinerjanya dengan metrik MAE (0.244), MSE (0.092), dan R-squared (0.893), menunjukkan tingkat akurasi yang cukup baik dalam memprediksi harga. Selanjutnya, model yang berhasil dikembangkan diimplementasikan ke dalam aplikasi web interaktif berbasis Streamlit. Aplikasi ini memungkinkan pengguna untuk memasukkan tanggal dan secara langsung mendapatkan prediksi harga beras premium. Hasil penelitian menunjukkan bahwa Regresi Linier efektif dalam memprediksi harga beras premium, dan implementasi ke dalam aplikasi Streamlit berhasil menyediakan alat prediksi yang mudah diakses. Meskipun demikian, penelitian lanjutan dapat berfokus pada peningkatan akurasi model dan eksplorasi algoritma Machine Learning lainnya untuk prediksi harga komodita
Mapping Library User Behavior Base On K-Means Clustering Of Ma’soem University Student: Pemetaan Perilaku Pengguna Perpustakaan Berbasis K-Means Pada Mahasiswa Universitas Ma’soem
Get PDFPerpustakaan memiliki peran strategis dalam mendukung kegiatan akademik di perguruan tinggi, namun rendahnya tingkat kunjungan dan peminjaman buku oleh mahasiswa menunjukkan adanya kesenjangan perilaku pengguna. Penelitian ini bertujuan untuk memetakan pola perilaku mahasiswa dalam memanfaatkan layanan perpustakaan dengan menerapkan algoritma K-Means Clustering terhadap data kunjungan dan peminjaman selama tiga bulan terakhir. Melalui pendekatan data mining, mahasiswa dikelompokkan ke dalam tiga klaster: aktif, cukup aktif, dan pasif, berdasarkan frekuensi kunjungan serta jumlah peminjaman buku. Proses analisis dilakukan menggunakan perangkat lunak RapidMiner dengan tahapan meliputi data preprocessing dan evaluasi model menggunakan indeks Davies-Bouldin. Hasil penelitian menunjukkan bahwa pendekatan klasterisasi efektif dalam mengidentifikasi segmentasi perilaku mahasiswa, yang selanjutnya dapat digunakan sebagai dasar dalam merumuskan kebijakan pengembangan layanan perpustakaan yang adaptif dan terarah. Penelitian ini memberikan kontribusi berupa kerangka kerja berbasis data
Design of Artificial Immune System - Models and Algorithms: Design of Artificial Immune System - Models and Algorithms
No full textArtificial Immune Systems (AIS) belong to a group of computational intelligence methods inspired by the working mechanisms of biological immune systems to solve various computational problems. Artificial Neural Networks (ANNs) themselves are often used in various fields such as anomaly detection, pattern recognition, cyber and network security, task scheduling, process optimization, and data analysis, with the application of various ANN algorithms. In the AIS approach, there are four basic algorithms that serve as the main foundation, namely the Negative Selection Algorithm (NSA), Artificial Immune Networks (aiNet), Clonal Selection Algorithm (CLONALG), and Dendritic Cell Algorithm (DCA). The problem that occurs at this time is that there is still a lack of papers that discuss the main basic algorithms in AIS, resulting in difficulties in developing new models of basic algorithms. Apart from that, many other aspects of the natural immune system have not been touched due to not yet understanding the basic algorithm of AIS. This paper aims to explain the main models and algorithms in AIS above so that in future research, new algorithms can be developed based on the basic algorithm as a reference. The results of this paper are a review of the main basic models and algorithms in AIS
From Natural Potential to Business Opportunity: Preparing Geotourism Based Entrepreneurship at Gunung Padang Site, Cianjur: Dari Potensi Alam ke Potensi Usaha: Menyiapkan Kewirausahaan Geowisata di Situs Gunung Padang Cianjur
Get PDFSitus Gunung Padang di Kabupaten Cianjur merupakan situs cagar budaya nasional yang menyimpan potensi besar sebagai destinasi geowisata berbasis konservasi dan edukasi. Namun, pengelolaan kawasan ini cenderung terjebak dalam pola mass tourism yang berorientasi pada kuantitas pengunjung, tanpa memperhatikan aspek pelestarian dan pemberdayaan masyarakat lokal. Artikel ini membahas inisiatif pengabdian masyarakat yang dilakukan oleh Universitas Bakrie dengan tujuan menyiapkan kewirausahaan berbasis geowisata di kalangan pemuda dan pelajar sekitar kawasan situs. Kegiatan ini merupakan program berkelanjutan yang diawali dengan survei pendahuluan pada April 2025 untuk memetakan potensi komunitas serta menyusun materi pelatihan. Metode pelaksanaan menggunakan pendekatan partisipatif berbasis komunitas, yang mencakup perancangan pelatihan kewirausahaan, pendampingan, serta penyusunan materi akademik yang relevan dengan prinsip geowisata. Hasil kegiatan awal menunjukkan tingginya antusiasme dan potensi keterlibatan generasi muda dalam praktik geo-entrepreneurship. Program ini diharapkan dapat mendukung tujuan pembangunan berkelanjutan (SDG 11), khususnya dalam mendorong pengelolaan kawasan wisata yang inklusif, edukatif, dan berkelanjutan
Multi-Classification of Pakcoy Plants using Machine Learning Methods with Smart Greenhouse Dataset
No full textThis research aims to design and implement a monitoring and classification system for Pakcoy (Brassica rapa L.) plant conditions based on the Internet of Things (IoT) and machine learning algorithms in the Smart Greenhouse of Universitas Islam Nusantara. This study represents one of the applications of IoT and machine learning technology advancements to improve efficiency and effectiveness in the agricultural sector. The developed system utilizes CO?, SHT30, BH1750, and DHT22 sensors to monitor environmental parameters in real-time, including temperature, humidity, light intensity, panel box temperature, and CO? concentration. The monitoring data are used as input for classifying plant conditions using five machine learning methods: Support Vector Machine (SVM), Random Forest, Decision Tree, Logistic Regression, and Multilayer Perceptron (MLP). The results show that the Random Forest algorithm achieves the best performance, with an accuracy of 84%, precision of 86%, recall of 87%, and F1-score of 86%. The implementation of this system serves as a concrete step toward enhancing the efficiency, sustainability, and modernization of hydroponic agriculture in Indonesi
Klasifikasi Status Mahasiswa Berisiko Drop Out Menggunakan Decision Tree C5.0 dengan Seleksi Fitur: Classification of Student Drop Out Risk Using Decision Tree C5.0 with Feature Selection
No full textPenelitian ini bertujuan untuk membandingkan performa algoritma Decision Tree C5.0 dengan dan tanpa seleksi fitur Mutual Information dalam klasifikasi mahasiswa berisiko drop out. Data yang digunakan mencakup mahasiswa S1 angkatan 2018–2023 dengan atribut performa akademik tiap semester serta aktivitas tugas akhir mahasiswa. Tahapan pengolahan data meliputi seleksi data, pembersihan, pelabelan, dan penanganan data tidak seimbang menggunakan Random Undersampling. Evaluasi model dilakukan menggunakan metrik akurasi, presisi, recall, dan f1-score. Hasil penelitian menunjukkan bahwa pada mahasiswa angkatan 2022–2023, model tanpa seleksi fitur menghasilkan akurasi sebesar 87,14%, dan meningkat menjadi 88,57% setelah penerapan seleksi fitur. Sementara itu, pada mahasiswa angkatan 2018–2021, akurasi model tanpa seleksi fitur mencapai 93,04% dan meningkat menjadi 94,30% dengan penerapan seleksi fitur. Faktor dominan yang memengaruhi klasifikasi berbeda pada masing-masing kelompok, di mana jumlah absen menjadi indikator utama pada mahasiswa angkatan 2022–2023, sedangkan durasi pengerjaan tugas akhir lebih berpengaruh pada mahasiswa angkatan 2018–2021
Analytical Hierarchy Process (AHP) Approach for CCTV Vendor Selection in Goods Procurement
No full textSelecting CCTV devices and vendors is a complex process because it involves many criteria, such as reputation/rating, price, distance, delivery time, and warranty period. Financing companies face difficulties in conducting these evaluations because the process is still manual, there are no standard criteria, and it requires many stages of approval that hinder the process. This study designs a web-based decision support system using the Analytic Hierarchy Process (AHP) method to assist in the objective selection of CCTV equipment. This system considers three main criteria: price, delivery time, and warranty period. The calculation results show that the system produces consistent recommendations, with a consistency ratio (CR) value below 0.1. The system also addresses zero data in alternatives by replacing zero values with small (non-zero) numbers to maintain the validity of AHP calculations. Repeated testing shows stable results even with different data. Thus, this system provides an efficient, transparent, and standardized solution in the CCTV device procurement process at financing companie
Perancangan Prototipe Sistem Bertenaga Mandiri Berbasis Pemanenan Energi Frekuensi Radio: Perancangan Prototipe Sistem Bertenaga Mandiri Berbasis Pemanenan Energi Frekuensi Radio
No full textElectronic devices are generally powered by a power source such as an AC socket or a DC power source such as a battery. Batteries have a limited lifespan, so it is necessary to recharge the battery if the battery is rechargeable. In this paper we present a design for an independent power system where the power does not come from an AC voltage socket, but the power is obtained from the results of radio frequency energy harvesting. The selected frequency is the frequency for cellular that emitted from the Cell Phone Operator's Base Transfer System. The Radio Frequency Energy Harvesting system catch RF signal use monopole antenna receiver, then the RF signal converted into DC voltage (RF to DC converter) by a series of charge pump that also serves as amplifier which designed to 5-stage villiard multiplier. This system integrated by boost regulator to raise the level of voltage and regulate constant voltage. The voltage result from system is 300 mV at 100 m from BTS and 3,9 V at 30 cm from mobile phone. Another test with a LED indicator and suplying a power for charging process to battery AAA rechargeable Ni-MH.
Keyword -cellular, energy harvesting, independent power system, monopole antenna, radio frequencyPerangkat elektronik umumnya ditenagai oleh sumber daya seperti stopkontak AC atau sumber daya DC seperti baterai. Baterai memiliki masa pakai yang terbatas, sehingga perlu diisi ulang jika baterai tersebut dapat diisi ulang. Dalam makalah ini, kami menyajikan perancangan sistem daya independen di mana daya tidak berasal dari stopkontak AC, melainkan daya diperoleh dari hasil pemanenan energi frekuensi radio. Frekuensi yang dipilih adalah frekuensi untuk seluler yang dipancarkan dari Sistem Transfer Basis Operator Telepon Seluler. Sistem Pemanenan Energi Frekuensi Radio menangkap sinyal RF menggunakan antena monopole, kemudian sinyal RF tersebut diubah menjadi tegangan DC (konverter RF ke DC) oleh rangkaian pompa pengisian daya yang juga berfungsi sebagai penguat yang dirancang untuk pengganda tegangan 5-tahap. Sistem ini terintegrasi dengan regulator penguat untuk menaikkan level tegangan dan mengatur tegangan konstan. Tegangan yang dihasilkan dari sistem adalah 300 mV pada jarak 100 m dari BTS dan 3,9 V pada jarak 30 cm dari telepon seluler. Pengujian lain dilakukan dengan indikator LED dan memasok daya untuk proses pengisian daya ke baterai AAA Ni-MH isi ulang.
Kata kunci: antena monopole, frekuensi radio, pemanenan energi, seluer, sistem daya independen
Community Empowerment of Perlis Village, West Brandan, Langkat through Mangrove Restoration and Clean Water Supply Activities : Pemberdayaan Masyarakat Desa Perlis, Brandan Barat, Langkat melalui Kegiatan Pemulihan Mangrove dan Penyediaan Air Bersih
No full textThe mangrove forests of Perlis Village, West Brandan, Langkat have suffered degradation due to illegal logging, agricultural conversion, and seawater intrusion, causing severe ecological and socioeconomic challenges. This study aimed to restore the mangrove ecosystem and improve access to clean water, addressing the community's urgent needs while supporting sustainable development goals. The project significantly enhanced the livelihoods of local fishermen, provided reliable clean water access, and fostered environmental stewardship. By combining ecological restoration with infrastructure development, it offered a replicable model for addressing environmental and social challenges in coastal communities. A participatory approach was employed, involving community engagement in planning, capacity-building workshops, and ecological restoration activities. Mangrove saplings were planted, and a clean water supply system was established through collaborative efforts between community members and external stakeholders. The project successfully restored 5 hectares of mangroves, increasing biodiversity and reducing coastal erosion. The survival rate of planted saplings was 85%. The clean water system expanded access from 40% to 85% of households, significantly reducing waterborne diseases and the time spent fetching water, especially for women and children. The initiative demonstrated that integrated community-driven interventions can address complex environmental and social issues effectively. The outcomes align with the study's objectives, providing valuable insights into sustainable development practices and the empowerment of coastal communities.
 
Educational Movement Against Hypertension and Diabetes (Gemes) in Prolanis Members: Gerakan Edukasi Mencegah Hipertensi dan Diabetes (Gemes) pada Anggota Prolanis
No full textCoverage of the Minimum Service Standards (MSS) for hypertension and diabetes in Puskesmas Air Putih is still low, reaching only 11.31% and 26%, respectively, while the fulfillment of the quality of service for each type of basic service in the health MSS must reach 100%. The GEMES (Educational Movement to Prevent Hypertension and Diabetes) program is a public health intervention that aims to increase public knowledge, especially prolanis members, about hypertension and diabetes mellitus. The method used is counseling with a series of pre-test activities, distribution of leaflet media, presentation of material and questions and answers, post-test, and reaction. Based on the results of the program evaluation through the pre-test and post-test tested using the Wilcoxon Test, the p.value was (0.002 <0.05), then H0 was rejected so it was concluded that there was an increase in knowledge in participants before and after receiving material on hypertension and diabetes mellitus. To support the sustainability of the program, it is recommended that health counseling be carried out regularly with more interesting material so that information is broader and the program is more effective.