AI Fusion with Adaptive Filter for Non-stationary Data in Parameter Prediction

In recent years, various fusion technologies have also been used to achieve performance improvement. When AI has been applied widely, one of the major challenges is the resistance of machine learning (ML) models to adapt to non-stationary data streams. Models have been increasingly required to update in a non-stationary environment [1]. In this paper, a fusion regression system is proposed to improve tracking capability for uncertainty and non-stationary data by combining machine learning models and adaptive filters.