Weather Forecasting using Deep Feed Forward Neural Network (DFFNN) and Fuzzy Outlier Removal
Weather forecasting is significant designed for people who facilitate to create additional informed daily decisions, and to keep out of danger. Precise Weather forecasting is becomes one of most difficult problems approximately the world. Unlike conventional methods, modern Weather forecasting consists of a combination of system models, examination. Data mining uses various technologies to forecast weather, predict rainfall, wind pressure, humidity, etc. Classification in data mining varied the restriction to view the clear information. The prediction of weather should be precise and moreover the weather be supposed to be forecasted previous at least a month before which determination be cooperative designed for many applications like agriculture, military, etc. Artificial Neural Networks (ANNs) have been applied comprehensively to together regress and classify weather experience. So in this work proposed Deep Feed Forward Neural Network (DFFNN) is rendering accurate predictions with noisy datasets, there is currently not a significant amount of research focusing on whether DFFNN are capable of producing accurate forecasts of relevant weather variables from small-scale, imperfect datasets. In addition the proposed work shows the outliers in the dataset is also removed by using fuzzy technique during the classification task. There is no important quantity of investigate focusing on the forecasting performance of Neural Networks realistic to weather datasets with the purpose of have been temporally rolled-up from a base dataset.
Author Name: M. Nagaselvi and Dr.T. Deepa
Author Email: -
Phone Number: -
Keywords: Data Mining, Classification, Weather Forecasting, Classification Methods, Soft Computing, Artificial Neural Network (ANN), Deep Feed Forward Neural Network (DFFNN) and Fuzzy Outlier Removal