<![CDATA[Unpredictable weather changes in Medan City often cause significant impacts across various sectors such as transportation, health, and the economy. One of the most frequent consequences is flooding due to high and poorly predicted rainfall. Therefore, this study aims to develop a weather prediction model by combining Linear Regression and Fuzzy Logic methods to improve the accuracy of weather parameter predictions, such as temperature, humidity, and rainfall measured in oktas. Linear Regression is used to model the statistical relationships between weather variables based on historical data, while Fuzzy Logic is applied to handle the uncertainty in weather patterns that are often nonlinear and unpredictable. This study utilizes historical weather data obtained from the worldweatheronline.com website for the Medan City area. The data goes through several processing stages, including preprocessing to clean and normalize the data, modeling using Linear Regression to identify major trends in the dataset, and applying Fuzzy Logic to accommodate uncertainties in predictions. The developed model is then tested and validated to evaluate its performance compared to other conventional prediction methods. The model implementation is carried out using GNU Octave, which allows efficient numerical analysis and data modeling. The results show that the combination of Linear Regression and Fuzzy Logic improves the accuracy of weather predictions compared to using a single method. This model provides more stable and reliable prediction results, especially in handling extreme weather conditions that frequently occur in Medan City. With a more accurate prediction model, it is expected to assist the community, government, and policymakers in disaster mitigation, infrastructure planning, and strategic decision-making in facing dynamic weather changes]]>
