[This article belongs to Volume - 54, Issue - 08]
Gongcheng Kexue Yu Jishu/Advanced Engineering Science
Journal ID : AES-20-10-2022-361

Title : Spread Spectral Deep Neural Network for Recommending Right Season and Weather Forecasting For Agriculture Development

Abstract : Precision agriculture analysis is a development fact in big data mining for predicting weather data for future recommendation agricluture. Especially in agricultural development, the spatial data is more difficult to predict right information because of more dimension due to non-relation feature analysis leads classification prediction problems. To resolve this, we propose a Forecasting weather prediction model based On Mutual Invariance Feature Selection Model (MIFSM) with intent of Successive Weather Influence Rate (WSIR) depended feature prediction and classified with Subset Spread Spectral Deep Neural Network (S3-DNN). Further predicting right features based on relevant features estimation using successive weather influence rate (SWIF) is estimated. Initially the proposed system collects the geo spatial weather data and process into feature selection using Mutual In variance Feature selection model. The features gets estimated using Successive Weather Influence Rate (SWIF) to make mean weightage along with marginal values observed from dataset rainfall, temperature, humidity etc. the estimated weight is patterned using Spatial Harvest Successive Rate (SHSR) to make ordered ranking. Further the selected features is trained into Soft Max Logical Activation Function (SMLAF) to get tuned neural network Using Convolution Neural Network (CCN). The classifier get trained with SMLAF to process input features make categorize the data into recommend and non-recommend fields based on this weather which is for recommendation for agricultural resources. The proposed system produce best recommendation performance as well than previous system.