Sampling-Free Probabilistic Deep State-Space Models
This work addresses the challenge of efficient inference in complex dynamical systems for researchers and practitioners in machine learning, representing an incremental improvement by introducing a deterministic method to an existing probabilistic framework.
The paper tackled the problem of inference in Probabilistic Deep State-Space Models (ProDSSMs) for dynamical systems with unknown parametric forms, proposing the first deterministic inference algorithm that achieves efficient approximations for training and testing, with results showing a superior balance between predictive performance and computational budget.
Many real-world dynamical systems can be described as State-Space Models (SSMs). In this formulation, each observation is emitted by a latent state, which follows first-order Markovian dynamics. A Probabilistic Deep SSM (ProDSSM) generalizes this framework to dynamical systems of unknown parametric form, where the transition and emission models are described by neural networks with uncertain weights. In this work, we propose the first deterministic inference algorithm for models of this type. Our framework allows efficient approximations for training and testing. We demonstrate in our experiments that our new method can be employed for a variety of tasks and enjoys a superior balance between predictive performance and computational budget.