Big Data Analytics · Machine Learning · ODE-based Mechanistic Modelling · Statistical Modelling · Applied AI Agents
My research focuses on big data analytics, machine learning, dynamic mechanistic modelling, statistical modelling, and AI agent applications, particularly in animal production and feed manufacturing. My work involves integrating and analyzing biological and production data, developing predictive models, simulating animal growth and metabolic processes, and supporting data-driven decision-making. By combining data analytics, mathematical modelling, and artificial intelligence, I aim to improve production efficiency, optimize feed manufacturing processes, and support the sustainability of animal production systems.
Data Integration Feature Engineering Data Imputation Anomaly Detection Machine Learning ODE-based Mechanistic Modelling Mixed-Effects Models Meta-Analysis Parameter Estimation Applied Agentic AI Python SAS Julia R
Developing a compartmental ODE-based mechanistic model to simulate growth and metabolic processes, including protein and fat turnover, in modern commercial turkey genetic lines. Collaboration with Hendrix Genetics. Manuscript in preparation.
Methods: ODE-based Mechanistic Modelling Parameter Estimation Sensitivity Analysis Julia
Developed machine learning and statistical models to predict Pellet Durability Index (PDI). Integrated nutrition factors, manufacturing parameters, and environmental variables to built robust prediction systems to support real-time decision-making at the mill level. Collaborated with Trouw Nutrition Canada, Molesworth Feed Supply, MasterFeeds, and Jones Feed Mills.
Methods: Data Integration Machine Learning Mixed-Effects Models Regression Feature Engineering Python SAS
This project applies machine learning to large-scale production data collected from precision-fed broiler breeders. The research focuses on predicting oviposition events and developing automated methods for detecting and cleaning anomalous body weight records, with the aim of improving data quality and supporting precision management of broiler breeders.
Methods: Machine Learning Classification Python
A Python library for Impact Range Assessment (IRA), supporting both single and repeated analyses to quantify the influence of input variables on regression model predictions and improve model interpretability.
Explainable AI Machine Learning Sensitivity Analysis Model Interpretability Regression Python
An interactive web application developed and deployed using machine learning and statistical modelling approaches to predict Pellet Durability Index (PDI) in commercial feed mills, supporting feed quality assessment and production process optimization.
Model Deployment Feed Manufacturing Pellet Quality Web App Python
This project aims to develop a large language model (LLM)-based framework for automated management of predictive models in feed mills. By integrating rule-based workflows with AI agents, the framework explores automated model monitoring, performance evaluation, and updating to support long-term model maintenance.
Large Language Model (LLM) Applied Agentic AI Model Updating Model Deployment Feed Manufacturing Pellet Quality Python
A lightweight cross-language toolkit for evaluating regression model predictions, with consistent Python (regreval) and Julia (RegrEval) implementations. It provides prediction-error metrics, Bibby–Toutenburg MSPE decomposition, Lin’s concordance correlation coefficient, etc.
Regression Evaluation Metrics MSPE Decomposition Concordance Correlation Coefficient Python Julia
🏫 University Profile: University of Guelph — Department of Animal Biosciences — Jihao You
🔗 LinkedIn: linkedin.com/in/jihao-y-1128b7193
📧 Email: jyou03@uoguelph.ca