Smart agriculture with big data

Abstract

This paper presents the importance of smart agriculture, which is the revolution in traditional farming using new technologies such as big data, Hadoop, R programming and cloud computing. This paper covers the major areas of agriculture from seed to the marketing level. This schema contains mainly three layers. In the First Layer that is data collection layer, collecting data using sensors related to weather and environmental changes, identification of soil types, crop identification ,yield tracking, water availability in the soil . Not only sensor data, Here we will take Historical data i.e. past data, by this prediction will be easy. All this data will be stored in HDFS i.e. Hadoop distributed file system .In the second layer that is Analysis layer, the collected information is analysed using data mining algorithms. The most important part in data analysis is data acquisition; we extract knowledge from sensor data in real time and historical data. In the third layer i.e. data prediction layer predictions obtained from the analysis based on that we will get the answers for the questions what to plant, when to plant and where to plant for this purpose we use data mining algorithms. The prediction discovers relationship between independent variables and relationship between dependent and independent variables. Many classification and regression algorithms are available for data prediction. Some of them are Decision Trees, Artificial neural networks, Support Vector Machine, Bayesian classification and Regression and K-means clustering. Among these we are using Decision Trees and K means clustering for prediction.