A feature based neural network model for weather forecasting

Abstract

TOPIC: NN model to weather forecasting TOOL: NN . PREDICTION GOAL: tmax, tmin and relative humidity INPUT: features extracted in different periods as well as from the weather parameter time-series itself from a particular station. NN ARCH: FF + BP. RESULTS: very encouraging (can make predictions with high degree of accuracy). EXTRA: can be adapted for larger geographical areas.