Publications
My publications span causal AI, forecasting, and machine learning.
Accepted
Causal Story-Guided Agentic Causal Discovery with Small Language Models
Brandon Mossop and Salimur Choudhury.
Accepted to the 4th International Conference on Foundation and Large Language Models (FLLM 2026).
This work extends agentic causal discovery with explicit causal-story reasoning and evaluates the approach on engineering datasets.
2026
Regime-aware causal Bayesian forecasting for non-stationary time series
Brandon Mossop and Salimur Choudhury.
PLOS ONE, 21(8), e0356474, 2026.
A regime-aware forecasting framework that combines temporal causal discovery, Bayesian structural models, predictive regime classification, and posterior predictive uncertainty.
Improving machine learning forecasting explainability and performance with causal feature selection for infectious disease spread
Brandon Mossop and Salimur Choudhury.
INFOR: Information Systems and Operational Research, 2026.
This work investigates causal feature selection for infectious-disease forecasting, with emphasis on interpretability and predictive performance.
2025
Agent-guided Causal Discovery with a Small Language Model
Brandon Mossop and Salimur Choudhury.
2025 3rd International Conference on Foundation and Large Language Models (FLLM), pp. 693–698, 2025.
An agent-guided causal discovery framework combining conditional-independence testing, contextual orientation, edge augmentation, and critique with a small language model.
2023
Forecasting Hospital Mental-Health Admissions with a Novel Hybrid Deep Learning Architecture
Brandon Mossop, Ramprasad Bismil, and Quazi Abidur Rahman.
2023 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 4093–4100, 2023.
A deep-learning forecasting approach for hospital mental-health admissions using multivariate time-series data.
Infectious Disease Forecasting using Multivariate Incomplete Time-series: A Hybrid Architecture with Stacked Dilated Causal Convolutions
Brandon Mossop and Quazi Abidur Rahman.
2023 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 4220–4227, 2023.
A hybrid deep-learning architecture for forecasting infectious-disease time series with incomplete multivariate observations.