WEATHER PREDICTION ANALYTICS USING MAPREDUCE-BASED LOGISTIC REGRESSION

Abstract

In the modern world, the prediction of weather becomes a challenging task. The weather prediction system is very useful and important for our agriculture. Weather forecasting is important for investigating of many business and decreases crop damage. Agriculture is the vital role for our country’s business and most of people are depend on developing activities. Regression is one of the main methods used in weather data prediction. Multinomial logistic regression is the important role of the system and forecasting of weather data based on temperature, humidity and wind. This system is applied to predict the weather data for a given location with multinomial logistic regression in order to obtain the desired prediction. The system of weather data prediction for a given location is Hinthada Region. It applies multinomial logistic regression and Map Reduce platform. In the system uses multinomial logistic regression to calculate the model. The various formats of weather datasets store in Hadoop Distributed File System and to obtain the optimum result the MapReduce Algorithm is used. This system predicts weather forecasting of Hinthada Region. Weather forecasting is one of the most important task for farmers in Hinthada Region and they decide their agriculture. Weather prediction helps agriculturist for decision making for their crop. The system helps farmers to use effective approach for weather prediction. Hinthada Region’s economy is highly dependent on its agricultural products. The system helps the agriculturist to get the awareness of their business and income.