The study explores the application of machine learning techniques to detect greedy nodes within a LoRaWAN network using real-time data. A detection methodology, is used to identify the greedy nodes in the network and, transforming it into a classification problem aimed at distinguishing between standard and greedy nodes. The methodology involves extracting transmission parameters from the dataset and computing the energy consumption of the data. By analyzing these factors, the model aims to discern whether nodes are operating optimally or employing greedy settings for data sensing across different applications. This study deals with the framework for detecting the nodes by predicting the energy consumption of each nodes distributed locally in a network.