A Method for Weather Forecasting Using Machine Learning

Agriculture is a sector that plays a crucial role in the economies of many countries around the globe, like India where it contributes 16% of the total economy. Weather forecasting is one of the challenges faced by this sector, due to its dynamic and turbulent nature, the statistical methods fail to provide forecasting at an accurate precision. This paper aims to develop an accurate way to predict the temperature forecast using machine learning techniques especially using Long Short Term memory networks (LSTM). Despite the advances made, there are still significant obstacles to overcome in expanding the use of weather forecasts in the agricultural sector due to the dynamics in climate changes. These include the need for improved model accuracy, quantitative evidence of the utility of climate predictions as instruments for agricultural risk management and addressing major chances of disease incidence which are usually seasonal and depends on parameters like temperature and rainfall. The goal of this study is to forecast parameters that could help farmers to make an informed decision so as to reduce the losses by taking required proactive measures. This paper provides a detailed analysis of weather forecasting techniques and explores future research goals in this field.