Approximate Probabilistic Inference for Time-Series Data: A Robust Latent Gaussian Model with Temporal Awareness

Anton Johansson, Arunselvan Ramaswamy

2025

Abstract

The development of robust generative models for highly varied non-stationary time-series data is a complex and important problem. Traditional models for time-series data prediction, such as Long Short-Term Memory (LSTM), are inefficient and generalize poorly as they cannot capture complex temporal relationships. In this paper, we present a probabilistic generative model that can be trained to capture complex temporal information, and that is robust to data errors. We call it Time Deep Latent Gaussian Model (tDLGM). Its novel architecture is an extension of the popular Deep Latent Gaussian Model (DLGM). Our model is trained to minimize a novel regularized version of the free energy loss function (an upper bound for the negative log loss). Our regularizer, which accounts for data trends, facilitates robustness to data errors that arise from additive noise. Experiments conducted show that tDLGM is able to reconstruct and generate complex time-series data. Further, the prediction error does not increase in the presence of additive Gaussian noise.

Paper Citation

in Harvard Style

Johansson A. and Ramaswamy A. (2025). Approximate Probabilistic Inference for Time-Series Data: A Robust Latent Gaussian Model with Temporal Awareness. In Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART; ISBN 978-989-758-737-5, SciTePress, pages 310-321. DOI: 10.5220/0013154800003890

in Bibtex Style

@conference{icaart25,
author={Anton Johansson and Arunselvan Ramaswamy},
title={Approximate Probabilistic Inference for Time-Series Data: A Robust Latent Gaussian Model with Temporal Awareness},
booktitle={Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART},
year={2025},
pages={310-321},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013154800003890},
isbn={978-989-758-737-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART
TI - Approximate Probabilistic Inference for Time-Series Data: A Robust Latent Gaussian Model with Temporal Awareness
SN - 978-989-758-737-5
AU - Johansson A.
AU - Ramaswamy A.
PY - 2025
SP - 310
EP - 321
DO - 10.5220/0013154800003890
PB - SciTePress