You have no items in your shopping cart.
A crucial part of life is predicting the weather. It is a scientific discovery that affects human existence. Various issues including disaster management, farming, ship and aircraft navigation, hydroelectric plant operation, sports and entertainment events, and even everyday home activities are impacted by accurate weather forecasts. Thus, this study developed a multiple linear regression model to predict daily temperature using historical meteorological data. The obtained data was analyzed using Python, a high-level, general-purpose programming language with code readability design. The model incorporates five key variables: maximum and minimum temperature, wind speed, wind direction, and humidity. The results show a strong positive correlation between temperature and maximum temperature and a significant inverse relationship between temperature and humidity. Evaluation metrics like R-squared, Mean Squared Error (MSE), and Root Mean Squared Error (RMSE) indicated a good fit and an average prediction error of approximately 1.35 degrees Fahrenheit. The k-fold cross-validation results suggest that the model generalizes well to unseen data. Applying the model to a new day's weather data resulted in a predicted temperature remarkably close to the actual value, highlighting its potential for real-world application. A major limitation of the study is the limited access to free weather data as most data sources require some form or payment before accessing them. The study recommended that the government and concerned stakeholders at all levels should ensure there is free access to data for research and academic purposes. This will go a long way to boost research and development in the nation and individual states. Additionally, further study should be conducted to identify and include weather variables that might influence temperature to enhance the model’s predictive power.
