Long Short-Term Memory and KNN based Approach for Smart Irrigation System using IoT

The inefficient use of water in common irrigation practices leads to significant water loss. In order to save water and speed up the process, an intelligent irrigation system is required. With the development of machine learning and the Internet of Things, an intelligent system may now be built to perform this task automatically with minimal human intervention. This study showcases an LSTM-RNN-trained recommendation system for enabling effective water use with minimal farmer interaction, this is all made possible through the Internet of Things. Internet of Things (IoT) sensors are deployed all around the farm to gather information about the soil and surrounding environment. The data is then stored and processed using ML methods on a cloud server, and the farmer will be provided with irrigation recommendations based on the findings. Preprocessing, Feature Selection, and Model Training are the three main components of the proposed system. Preprocessing with an M3 filter is employed. RFE is used to perform feature selection. In the end, it's put to use in LSTM - RNN model training. When compared to LSTM and RNN models, the proposed method performs better.