Weather forecasts help improve farming, transportation, crisis management and daily planning activities. With traditional forecasting, detailed mathematical models are required and a large amount of computing power is necessary. Over the past few years, using machine learning has improved both the accuracy and speed of weather forecasts. We study the Random Forest algorithm in this paper, as it attempts to forecast weather accurately. Weather last year information like temperature, humidity, wind speed etc. the proposed model is trained to predict the using atmospheric pressure, are used to do this. The forecast of the day. By employing such common tools as accuracy, precision and mean squared error of Random Forest is examined and analyzed in comparison with other machine learning models. The outcomes show that using Random Forest greatly improves the accuracy and depends robustly on forecasting short-term weather patterns. According to this research, Random Forest shows promise as a way to design trustworthy and scalable forecast systems for weather.