Advanced Machine Learning Techniques for Data Prediction in WSNs
Sowjanya Bojja, Akula Rajitha, K Aravinda, Amandeep Nagpal, Ravi Kalra, Usama Kadem Radi · 2024
Modern data-receiving and monitoring systems use wireless sensor networks (WSNs); however, sensor data prediction can be difficult. A new machine learning system named Adaptive Sensor Ensemble (ASE) is introduced in this work. It improves WSN data prediction by combining the best of LSTM, RF, and AdaBoost. Our technique uses WSN sensor data’s unique properties to produce more accurate, energy-efficient, and real-time forecasts. LSTM networks detect temporal correlations, RF networks handle diverse sensor data, and AdaBoost networks adaptively calculate sensor predictions. The ASE algorithm adapts to changing sensor qualities and data patterns to make accurate predictions while keeping energy and computing power in mind in real time. ASE, revealed by “ML Net-WiSe,” outperforms typical approaches in accurate predictions, energy efficiency, real-time data processing, adaptability, scalability, and privacy. This indicates our superior WSN data forecast technology works.