A neural network approach for disease forecasting in grapes using weather parameters

Meteorological parameters such as temperature, rainfall and humidity are important for agricultural systems. Weather fluctuations and improper cultivation methods lead to loss of crop productivity. Therefore weather forecasting is of substantial importance to overcome these problems. Inclusion of predicted weather data could provide valuable and timely information for evaluation of various crop management techniques, to avoid potential losses thereby increasing crop production and income. The proposed work intends to predict the weather using a modified k-Nearest Neighbor (NN) approach, and Feed Forward Neural Network, and then utilize parameters such as humidity and temperature to predict the disease outbreaks in grapes. Such predictions would warn the growers of expected significant developments in grape disease via email or text messages. This work can be further extended to provide other quality information such as quantity of pesticide to be used if found necessary.