PRRS Outbreak Prediction via Deep Switching Auto-Regressive Factorization Modeling

• DOI: 10.48550/arxiv.2110.03147
• Publication Date: 2021-01-01
• Authors: M. Shamsabardeh; Bahar Azari; Beatriz Martínez‐...
• Keywords: 2. Zero hunger; 0403 veterinary science; FOS: Computer and information sciences; Computer Science - Machine Learning; 0202 electrical engineering, electronic engineering, information engineering; 04 agricultural and veterinary sciences; 02 engineering and technology; 3. Good health; Machine Learning (cs.LG)

We propose an epidemic analysis framework for the outbreak prediction in livestock industry, focusing on study of most costly and viral infectious disease swine industry -- PRRS virus. Using this framework, we can predict all farms a production system by capturing spatio-temporal dynamics infection transmission based intra-farm pig-level virus dynamics, inter-farm pig shipment network. simulate network SEIR model using statistics extracted from real data provided industry. develop hierarchical factorized deep generative that approximates high dimensional product between time-dependent weights spatially dependent low factors to perform per farm time series prediction. The results demonstrate ability forecasting spread progression with average error NRMSE = 2.5\%.