Article Per Year (5 Year)
p-Index From 2021 - 2026
0.61
P-Index
This Author published in this journals
All Journal Applied Technology and Computing Science Journal Journal of Applied Mathematics and Modelling
Radhiatul Husna
Andalas University
Computer Science & IT Control & Systems Engineering Engineering Environmental Science Mathematics Social Sciences

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

 1 
Development of a Rice Leaf Disease Detection Application Using Python-Based Computer Vision and YOLO Fajar Maulana; Yomei Hendra; Guswita Helmi; Radhiatul Husna; Amma Liesvarastranta Haz; Evianita Dewi Fajrianti
TEKNOLOGI DITERAPKAN DAN JURNAL SAINS KOMPUTER Vol 8 No 2 (2025): December
Publisher : Universitas Nahdlatul Ulama Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33086/atcsj.v8i2.8486

Abstract

Rice is one of Indonesia’s primary agricultural commodities and is highly vulnerable to various leaf diseases, including blast, blight, brown spot, and tungro, which can significantly reduce crop productivity. To address this issue, an automated and accurate detection system is needed to assist farmers in identifying rice leaf diseases at an early stage. This study aims to develop a rice leaf disease detection application using computer vision technology based on Python and the YOLO (You Only Look Once) algorithm. The research methodology consisted of several stages: problem identification, data acquisition, data exploration, model development, evaluation, and deployment. The dataset was obtained from Roboflow and comprised five classes: blast, blight, brown spot, healthy, and tungro. The YOLO model was trained using Google Colab with optimized parameters to enhance detection performance. Experimental results demonstrate that the proposed model achieved an accuracy of 95% and a mean Average Precision (mAP) of 95%, indicating strong performance in detecting and classifying rice leaf diseases. The system was implemented as a web-based application using Flask and Bootstrap, allowing users to upload images of rice leaves and obtain real-time detection results. This application enables farmers to identify plant diseases quickly and accurately, facilitating timely and effective intervention to minimize crop losses.
Simulation and Optimization of Review Intervals Based on the Mathematical Model of the Ebbinghaus Forgetting Curve Radhiatul Husna; Nabila Gusti Rohima; Alwan Ronan; Mursyid Nur Fahmi; Amma Liesvarastranta Haz; Evianita Dewi Fajrianti
TEKNOLOGI DITERAPKAN DAN JURNAL SAINS KOMPUTER Vol 8 No 2 (2025): December
Publisher : Universitas Nahdlatul Ulama Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33086/atcsj.v8i2.8491

Abstract

Humans naturally experience memory decay over time due to the brain’s limited capacity, a phenomenon first systematically quantified by Hermann Ebbinghaus in 1885 through the forgetting curve, which illustrates the exponential decline of retention in the absence of reinforcement. This curve demonstrates that newly acquired information fades rapidly unless reviewed, negatively affecting educational outcomes as students struggle to retain knowledge in the long term. Spaced repetition, involving scheduled review sessions, has emerged as an effective strategy to counteract forgetting; however, optimal review intervals are often determined intuitively rather than derived mathematically. This study aims to model memory retention dynamics using an extended Ebbinghaus forgetting curve formulated through a learning and forgetting differential equation model, estimate parameters from empirical data, and optimize review intervals to enhance long-term retention. Data were collected through questionnaires from 15 students in the 2023 Mathematics Study Program at Andalas University enrolled in Real Analysis I, yielding parameter estimates of learning rate (α = 0.70), forgetting rate (λ = 0.30), initial knowledge level (K(0) = 100%), and maximum knowledge capacity (K_max = 100%). The model was solved analytically, and numerical simulations compared three strategies: no review, random review, and optimal review at an 80% retention threshold. The optimal review time was found to be t = 1.098 days (approximately 26 hours and 32 minutes), corresponding to the point at which retention declines to 80%. Simulations showed that no review leads to near-zero retention over time, random review produces inconsistent improvements, and the optimal review strategy maintains retention above 80% efficiently. Overall, the mathematically derived optimal review strategy significantly outperforms alternative approaches, providing a personalized, evidence-based method to improve learning efficiency and long-term memory stability while demonstrating the value of integrating psychological memory theory with mathematical optimization for practical educational applications.
Graph Coloring Application in Course Scheduling Using the Welch-Powell Algorithm (Case Study: Mathematics Study Program, Dharma Andalas University) Salsabilla Mirza; Iswan Rina; Radhiatul Husna
Journal of Applied Mathematics and Modelling Vol. 2 No. 1 (2026): Journal of Applied Mathematics and Modelling
Publisher : CIB Nusantara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.64570/jamm.v2i1.58

Abstract

Course scheduling is a crucial and routine academic activity carried out by universities prior to the commencement of a new academic year, requiring precise arrangements to avoid timing conflicts for both lecturers and students. The course scheduling for the odd semester of the 2024/2025 academic year at the Mathematics Study Program of Dharma Andalas University (UNIDHA) faces high complexity in manually allocating various combinations of lecturers, classes, and courses. This study aims to optimize the scheduling system by utilizing the concept of graph coloring, specifically employing the Welch-Powell Algorithm. The research method involves transforming the lecturer's teaching assignment data into an undirected graph model. Course variations are modeled as 21 vertices, while potential scheduling conflicts due to identical lecturers or student groups are represented as edges. The Welch-Powell Algorithm is then applied by sorting the vertices based on the largest degree ordering to color them progressively. The results indicate that this graph representation successfully yields a chromatic number of x(G) = 6. Consequently, from a total of 9 lecturers teaching 21 courses, the entire academic schedule is optimally distributed into 6 distinct color groups without any conflicts. These color groups are successfully implemented into 6 academic working days (Monday to Saturday) with precise time-slot allocations corresponding to the respective credit semester units (SKS). The study concludes that the application of graph coloring using the Welch-Powell Algorithm is highly effective, objective, and systematic in resolving course scheduling conflicts within the Mathematics Study Program at UNIDHA.
Co-Authors Alwan Ronan Amma Liesvarastranta Haz Amma Liesvarastranta Haz Evianita Dewi Fajrianti Evianita Dewi Fajrianti Fajar Maulana Guswita Helmi Iswan Rina Mursyid Nur Fahmi Nabila Gusti Rohima Salsabilla Mirza Yomei Hendra