RegFlow: Probabilistic Flow-based Regression for Future Prediction
This work provides a more flexible and robust method for future prediction in non-deterministic systems, which is important for researchers and practitioners working with complex behavioral modeling, such as human behavior.
This paper addresses the challenge of predicting future states in complex, non-deterministic scenarios by introducing RegFlow, a probabilistic framework that models future predictions without strong assumptions about modality or underlying probability distributions. It leverages a hypernetwork architecture and trains a continuous normalizing flow model, achieving state-of-the-art results on several benchmark datasets.
Predicting future states or actions of a given system remains a fundamental, yet unsolved challenge of intelligence, especially in the scope of complex and non-deterministic scenarios, such as modeling behavior of humans. Existing approaches provide results under strong assumptions concerning unimodality of future states, or, at best, assuming specific probability distributions that often poorly fit to real-life conditions. In this work we introduce a robust and flexible probabilistic framework that allows to model future predictions with virtually no constrains regarding the modality or underlying probability distribution. To achieve this goal, we leverage a hypernetwork architecture and train a continuous normalizing flow model. The resulting method dubbed RegFlow achieves state-of-the-art results on several benchmark datasets, outperforming competing approaches by a significant margin.