Time Series Numerical Association Rule Mining for assisting Smart Agriculture

Smart agriculture is a modern paradigm that incorporates modern technologies ranging from cloud computing, big data, Internet of Things, or data mining, among others. The final objective of smart agriculture is to assist the classical farming process in automating some essential steps, and making agriculture more sustainable. In agriculture, different time-series data allow the application of apply machine learning methods in order to solve different problems. Currently, most of the techniques were developed for classification or prediction tasks. On the contrary, this paper sheds light on the potential of numerical association rule mining for time series data. In line with this, a novel method for mining time series numerical association rules is proposed on time-series data collected by monitoring parameters detected during a growth of plants. Experiments reveal that time series numerical association rule mining could be a suitable method for mining time-series data in agriculture and related areas.