كنفرانس بين المللی مهندسی صنايع و سيستم ها

صفحه اصلی / یازدهمین كنفرانس بين المللی مهندسی صنايع و سيستم ها

Cold Temperature Risk Prediction Using Machine Learning: Binary Classification of Minimum Daily Temperature in Sarakhs

نویسندگان :

Mohammad Farrokhpour¹ Mahdi Habibii² Amirhossein Babaeian³ Alireza Shadman⁴

1- دانشگاه فردوسی مشهد 2- دانشگاه فردوسی مشهد 3- دانشگاه فردوسی مشهد 4- دانشگاه فردوسی مشهد

کلمات کلیدی :

Minimum temperature،Machine learning،Frost prediction،Neural network،Supervised learning،classification

چکیده :

Abstract— Accurate forecasting of minimum daily temperatures is critical in preventing cold-related damages, particularly in agriculture. This study aims to classify the minimum temperature of the next day in Sarakhs, Iran, into two categories: below or above 4°C, using supervised machine learning algorithms. A dataset of meteorological observations was collected, including temperature, humidity, wind direction and speed, pressure, and solar radiation. The data was preprocessed, normalized, and labeled based on the 4°C threshold. Various classification algorithms such as Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), Support Vector Classifier (SVC), K-Nearest Neighbors (KNN), and Artificial Neural Networks (ANN) were evaluated. Model performance was assessed using accuracy, recall, precision, and F1-score metrics. Among the tested models, QDA and ANN demonstrated the highest performance, achieving F1-scores of 0.94 and 0.93, respectively. The results highlight the effectiveness of lightweight machine learning models in cold temperature risk classification and provide a foundation for future integration into early-warning systems for frost-prone regions.