Claim Missing Document
Check
Articles

IMPLEMENTASI PROFILE MATCHING PADA SISTEM PENCARI KERJA DISABILITAS Muhammad Dwi Ramadhianto; Syarifah Putri Agustini Alkadri; Barry Ceasar Octariadi
JIPI (Jurnal Ilmiah Penelitian dan Pembelajaran Informatika) Vol 10, No 4 (2025)
Publisher : STKIP PGRI Tulungagung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29100/jipi.v10i4.6686

Abstract

Penyandang disabilitas adalah seseorang yang mempunyai (menderita) sesuatu dan kecacatan adalah suatu keadaan (seperti penyakit atau cedera) yang membatasi atau merusak kemampuan fisik dan mental seseorang. Penelitian ini bertujuan untuk mengembangkan sarana penunjang pencari pekerjaan bagi penyandang disabilitas melalui website DISKA (Disabilitas Kerja), yang menggunakan metode Profile Matching. Sistem pendukung keputusan (SPK) ini mengintegrasikan aspek utama, aspek pendukung, dan aspek wawancara dalam penilaian kandidat. Berdasarkan hasil penelitian yang telah dilakukan selama pengembangan   Implementasi Profile Matching Pada Sistem Pencari Kerja Disabilitas. Aplikasi dirancang dan dibangun dengan metode profile matching digunakan untuk menentukan nilai yang terdapat pada core factor dan Secondary factor dan nilai aspek juga menetukan nilai akhir terhadap perhitungan dalam merekomendasi pekerjaan. Berdasarkan hasil dari aplikasi rekomendasi pekerjaan dengan metode profile matching dapat disimpulkan bahwa khauril anwar cocok dengan pekerjaan Barista dikarekanakan dekatnya hasil total antara pekerjaan dengan nilai 4.58 dan pelamar dengan nilai 4.67, ramadianti cocok dengan pekerjaan Art dikarekanakan hasil sama dengan pekerjaan Art dengan nilai 4.27 dan nilai pelamar yaitu 4.27, ahmad syahdani cocok dengan pekerjaan cleaning service karena nilai totalnya yaitu 4.455 melebihi nilai total dari cleaning service, famin nasrul cocok dengan pekerjaan cleaning service dikarenakan nilainya 4,36 melebihi nilai dari pekerjaan cleaning service karena itu cocok dengan pekerjaannya dan ananda dinda cocok dengan pekerjaan Pijat dikarenakan nilainya yaitu 4.215 melebihi nilai dari pijat oleh karena itu cocok dengan pekerjaan pijat.
Household intention to eliminate mosquito breeding sites following a single-session Geplakkin intervention Wahyu Naufal; Andri Dwi Hernawan; Elly Trisnawati; Barry Ceasar Octariadi
Jurnal Cakrawala Promkes Vol. 8 No. 2 (2026): August
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/jcp.v8i2.16846

Abstract

Dengue remains an important public health problem, and household elimination of mosquito breeding sites is central to prevention. Digital technologies may strengthen dengue prevention by integrating health information with vector surveillance, but evidence regarding their relationship with household behavioral intention remains limited. This study evaluated household intention to eliminate mosquito breeding sites following a single-session intervention using the Geplakkin vector surveillance application. A quantitative experiment with a posttest-only control-group design was conducted in Pontianak City, West Kalimantan, Indonesia, from July to October 2025. A total of 320 purposively recruited participants were randomly allocated in a 3:1 ratio to the intervention (n = 240) or control group (n = 80) using a computer-generated sequence. Both groups received routine dengue-prevention education; the intervention group additionally received 15–30 minutes of training on Geplakkin, followed by immediate application use and posttest assessment. Intention was assessed using a Theory of Planned Behavior-based questionnaire. Because scores were not normally distributed, groups were compared using the Mann–Whitney U test. The intervention group had a higher mean intention score than the control group (45.43 ± 4.68 vs 36.73 ± 5.50), with a mean difference of 8.70 points (p < 0.001; effect size = 0.65). Thus, exposure to the single-session Geplakkin intervention was associated with higher immediate intention to eliminate mosquito breeding sites than routine education alone. Longitudinal studies are needed to determine whether this difference is sustained and translates into household vector-control practices.
Penerapan Algoritma Random Forest untuk Rekomendasi Penerima Bantuan Pertanian CPCL Andri Santoso; Barry Ceasar Octariadi; Asrul Abdullah
Jurnal Teknologi Dan Sistem Informasi Bisnis Vol. 8 No. 3 (2026): Juli 2026
Publisher :  Prodi Sistem Informasi Universitas Dharma Andalas

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47233/jteksis.v8i3.189

Abstract

Manual determination of eligibility for Prospective Farmers and Prospective Locations (CPCL) assistance in swamp rice farming often leads to subjectivity, time-consuming processes, and misallocated aid. This study aims to construct a classification model to evaluate the eligibility of CPCL swamp rice assistance recipients in Kubu Raya Regency for 2024 using the Random Forest algorithm. The CPCL dataset underwent a comprehensive preprocessing pipeline, including data cleaning, missing value handling, categorical variable encoding, and feature selection. The model was trained using an 80:20 data split, tested on 26 test samples (7 "Ineligible" and 19 "Eligible"), and evaluated using Confusion Matrix metrics and Stratified 10-Fold Cross-Validation. The results demonstrate that the Random Forest model achieved an accuracy of 84.62% with a weighted average of 87% for precision, 85% for recall, and 82% for F1-score. For the "Eligible" class, the model achieved a recall of 1.00, precision of 0.83, and F1-score of 0.90. Meanwhile, for the "Ineligible" class, the model attained a precision of 1.00, recall of 0.43, and F1-score of 0.60. Feature importance analysis revealed that seed requirements, fertilizer allocation, pesticide supply, and land area were the primary factors influencing prediction outcomes.
KLASIFIKASI CITRA TANAMAN MERAMBAT LIAR BERKHASIAT OBAT MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORK Dini Dini; Barry Ceasar Octariadi; Sucipto Sucipto
HEALTHY : Jurnal Inovasi Riset Ilmu Kesehatan Vol. 5 No. 4 (2026)
Publisher : Pusat Pengembangan Pendidikan dan Penelitian Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51878/healthy.v5i4.14198

Abstract

Indonesia has high biodiversity of medicinal plants; however, the specific identification of wild climbing plants is often challenging due to similarities in leaf morphology among species, which may lead to errors in medicinal utilization. This study aims to develop an automatic identification system for five wild climbing medicinal plants, namely Betel, Brotowali, Kaduk, Water Pepper, and Binahong. The proposed method applies deep learning using a Convolutional Neural Network (CNN) algorithm with a transfer learning approach based on the MobileNetV2 architecture. The dataset consisted of 1,000 leaf images evenly distributed into five classes, followed by image preprocessing, data augmentation, and model training using the Adam optimizer with a learning rate of 0.0001. The evaluation results showed that the model achieved a validation accuracy of 95.31% during training and obtained an internal testing accuracy of 99.5% on previously unseen data. However, external testing using 25 images captured under real-world background conditions achieved an accuracy of 84%, indicating performance degradation due to variations in visual conditions outside the training dataset. The model successfully recognized most plant classes with high performance, with misclassifications mainly occurring between Kaduk and Betel due to their similar visual characteristics. The developed model was implemented into an interactive web-based interface using Streamlit as a supporting tool for medicinal plant identification. This study demonstrates that MobileNetV2 with a transfer learning approach is effective for classifying wild climbing medicinal plants using a limited dataset; however, increasing data diversity is still required to improve model generalization under real-world conditions. ABSTRAK Indonesia memiliki biodiversitas tanaman obat yang sangat tinggi, namun identifikasi spesifik terhadap tanaman merambat liar sering mengalami kendala akibat kemiripan morfologi daun antarspesies yang dapat menyebabkan kesalahan pemanfaatan. Penelitian ini bertujuan mengembangkan sistem identifikasi otomatis untuk lima jenis tanaman merambat liar berkhasiat obat, yaitu Sirih, Brotowali, Kaduk, Ketumpang Air, dan Binahong. Metode yang digunakan adalah deep learning dengan algoritma Convolutional Neural Network (CNN) melalui pendekatan transfer learning menggunakan arsitektur MobileNetV2. Dataset penelitian terdiri atas 1.000 citra daun yang terbagi secara seimbang ke dalam lima kelas, kemudian dilakukan tahap pra-pemrosesan, augmentasi data, dan pelatihan model menggunakan optimizer Adam dengan learning rate 0,0001. Hasil evaluasi menunjukkan bahwa model memperoleh akurasi validasi sebesar 95,31% selama proses pelatihan dan mencapai akurasi pengujian internal sebesar 99,5% pada data yang belum pernah digunakan sebelumnya. Namun, pengujian eksternal menggunakan 25 citra dengan kondisi latar belakang lingkungan nyata menghasilkan akurasi sebesar 84%, yang menunjukkan adanya penurunan performa akibat variasi kondisi visual di luar dataset pelatihan. Model mampu mengenali sebagian besar kelas tanaman dengan baik, dengan kesalahan klasifikasi terutama terjadi pada tanaman Kaduk dan Sirih yang memiliki kemiripan karakteristik visual. Model kemudian diimplementasikan dalam antarmuka web interaktif berbasis Streamlit sebagai media bantu identifikasi tanaman obat. Penelitian ini menunjukkan bahwa MobileNetV2 berbasis transfer learning efektif untuk klasifikasi citra tanaman merambat liar berkhasiat obat pada dataset terbatas, tetapi peningkatan keragaman data masih diperlukan untuk meningkatkan kemampuan generalisasi pada kondisi nyata.
Prediksi Status Gizi Pada Balita Menggunakan Metode Long ShortTerm Memory Hizrah Azhari; Barry Ceasar Octariadi; Rachmat Wahid Saleh Insani
Jurnal Publikasi Teknik Informatika Vol. 4 No. 1 (2025): Januari: Jurnal Publikasi Teknik Informatika
Publisher : Lembaga Pengembangan Kinerja Dosen

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55606/jupti.v4i1.5506

Abstract

The high prevalence of malnutrition among toddlers in Pontianak City, adversely affecting growth, cognitive function, and future human capital. Despite monitoring efforts, traditional methods remain slow and less accurate. The objective is to develop an early prediction system for toddler nutritional status using a Long Short-Term Memory (LSTM) model based on anthropometric data (weight, height, age, gender) from the Pontianak City Health Office. The methodology comprises an applied experimental quantitative design with the following stages: data collection and preprocessing (cleaning, normalization, train-test split), LSTM model architecture development, model training, and performance evaluation using accuracy, precision, recall, and F1-score metrics. The dataset includes 1,182 samples of toddlers aged 0–60 months. Training results yielded 82% accuracy, 98% precision, 82% recall, and an F1-score of 89% on the training set. On the test set, the model achieved 79% accuracy, 99% precision, 80% recall, and an F1-score of 89%, with a weighted average F1-score of 0.83 and a macro average F1-score of 0.71, indicating solid overall performance but highlighting the need for improvement in underrepresented classes. The study’s implications involve enhancing the effectiveness of toddler nutritional monitoring, enabling faster interventions by healthcare workers, and contributing academically to the development of AI-based applications in public health.
PREDIKSI OMZET PENJUALAN KIOS PANGAN MANDIRI BERDASARKAN FAKTOR PENJUALAN MENGGUNAKAN REGRESI LINEAR BERGANDA Dayang Nisa Anjani; Barry Ceasar Octariadi; Rachmat Wahid Saleh Insani
NUSANTARA : Jurnal Ilmu Pengetahuan Sosial Vol 13, No 10 (2026): NUSANTARA: Jurnal Ilmu Pengetahuan Sosial
Publisher : Universitas Muhammadiyah Tapanuli Selatan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31604/jips.v13i10.2026.2482-2488

Abstract

Kios Pangan Mandiri merupakan toko sembako yang dikelola pemerintah provinsiKalimantanBaratmelaluidinasketahananpanganuntukmenjagastabilitas harga dan ketersediaan bahan pokok bagi masyarakat. Meskipun memiliki tujuan sosial, kios ini tetap menjalankan aktivitas operasional layaknya sebuah usaha. Permasalahan yang dihadapi adalah kios masih mengalami kesulitan dalam memperkirakan omzet pada periode mendatang, sehingga berisiko menimbulkan ketidaktepatan dalam pengelolaan keuangan. Penelitian ini bertujuan untuk menganalisis dan mengimplementasikan metode regresi linear berganda dalam memprediksi omzet penjualan pada Kios Pangan Mandiri, serta mengukur tingkat akurasinya. Data yang digunakan adalah data historis transaksi harian kios pada periode November 2023 hingga April 2026 dengan empat variabel independen, yaitu waktu, total barang terjual, awal bulan, dan hari besar. Sistem prediksi dibangunmenggunakanbahasapemrogramanPythondenganframeworkStreamlit sebagai antarmuka berbasis web, sedangkan evaluasi tingkat akurasi model dilakukanmenggunakanmetode Mean Absolute Percentage Error(MAPE).Hasil penelitianmenunjukkanbahwametode regresi linearberganda berhasil diterapkan dalammembangunsistemprediksiomzetdengannilaiMAPEsebesar5.22%,yang berdasarkantabelinterpretasiMAPEtermasukdalamkategoriprediksisangatbaik (MAPE10%).Halinimenunjukkanbahwametoderegresilinearbergandalayak digunakan sebagai dasar perencanaan pengelolaan keuangan dan pengambilan keputusan strategis pada Kios Pangan Mandiri.