Prakasam et al. (2026) A Generative Model for Rainfall Prediction based on Variational Autoencoder (VAE) Using Time-Series Weather parameters

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Springer Link (Chiba Institute of Technology)
• Year: 2026
• Date: 2026-04-09
• Authors: S. Prakasam, S. Hariharan, P. Shanmugapriya
• DOI: 10.1051/itmconf/20268501002/pdf

Research Groups

Not specified in the paper.

Short Summary

This paper proposes a novel generative probabilistic rainfall prediction framework based on Variational Auto Encoders (VAE) that improves probabilistic accuracy and uncertainty calibration compared to traditional deterministic methods.

Objective

• To develop a novel generative probabilistic framework using Variational Auto Encoders (VAE) for reliable rainfall prediction that captures the inherent uncertainty in atmospheric processes.

Study Configuration

• Spatial Scale: Not explicitly specified, implied local/point-based.
• Temporal Scale: Daily

Methodology and Data

• Models used: Variational Auto Encoders (VAE)
• Data sources: Weather data

Main Results

• The VAE-based approach significantly improves probabilistic accuracy and uncertainty calibration for rainfall prediction compared to traditional deterministic methods.
• The proposed generative framework provides an interpretable latent representation of atmospheric states.

Contributions

• Introduction of a novel generative probabilistic rainfall prediction framework utilizing Variational Auto Encoders (VAE).
• Enhanced probabilistic accuracy and uncertainty calibration in rainfall prediction.
• Provision of an interpretable latent representation of atmospheric states for reliable prediction.

Funding

Not specified in the paper.

Citation

@article{Prakasam2026Generative,
  author = {Prakasam, S. and Hariharan, S. and Shanmugapriya, P.},
  title = {A Generative Model for Rainfall Prediction based on Variational Autoencoder (VAE) Using Time-Series Weather parameters},
  journal = {Springer Link (Chiba Institute of Technology)},
  year = {2026},
  doi = {10.1051/itmconf/20268501002/pdf},
  url = {https://doi.org/10.1051/itmconf/20268501002/pdf}
}