Team Leader
Das Kulangara Veettil
Texas A&M
Animal Science
vishnudasveettil@tamu.edu
Project Type
Research
Who Can Join
Masters Students, Undergraduate Students
Project Description
This project focuses on improving economic decision making in the stocker/backgrounding segment of the U.S. beef value chain, particularly in Texas. The stocker phase where weaned calves are grown on forage or roughage based diets before entering feedlots is economically complex, influenced by seasonal prices, health management, forage availability, and market volatility. Despite its importance, this stage often receives less modeling attention than cow-calf or feedlot operations.
Our research applies machine learning and time series forecasting techniques to build a comprehensive economic modeling pipeline tailored to stocker operations. Using Python, the project combines univariate time-series models (e.g., ARIMA, SARIMA, SARIMAX) and multivariate regression and ensemble models (e.g., Random Forest, AdaBoost, Support Vector Regression). Data will be collected through web scraping and real-time agricultural databases, as well as simulated via synthetic generation for underrepresented variables.
Key indicators costs, revenue, daily gain efficiency, and gross margin will be predicted and analyzed. A major component of the project involves incorporating probabilistic modeling to capture uncertainty and provide robust, risk-informed forecasts for producers and stakeholders.
Ultimately, this research aims to deliver a scalable and replicable model that can forecast profitability, inform operational planning, and support policy or investment decisions in stocker production systems.
Team Needs
We are seeking undergraduate students with a strong interest in applying data science and machine learning to real world agricultural challenges.
Preferred Skills and Backgrounds:
• Python proficiency (required), with experience using libraries like Pandas, NumPy, Matplotlib, Statsmodels, and Scikit-learn
• Familiarity with time-series analysis, machine learning models.
• Comfort with web scraping, or database management
• Strong communication skills and a collaborative attitude
Expectations:
• Commitment to weekly meetings and regular progress updates
• Willingness to learn new tools, review related literature, and contribute to group documentation
• Respect for interdisciplinary teamwork and openness to integrating economic, biological, and computational perspectives
No prior agricultural experience is required. Students from Animal Science, Computer Science, Statistics, Agricultural Economics, or any quantitative field are encouraged to apply.
Special Opportunities
This project offers undergraduates a chance to work at the intersection of agriculture, data science, and economics.
As a team member, you will:
• Gain hands-on experience in applying machine learning to real world beef industry data
• Work with time-series and economic datasets that reflect actual cattle production conditions in Texas
• Build skills in forecasting, regression, and uncertainty quantification
• Present findings at campus conferences, research symposia, or undergraduate research showcases
• Have the opportunity to earn co-authorship on peer-reviewed publications
• Contribute to the development of a scalable decision-support tool that may be used by ranchers, economists, or policymakers
• Collaborate in a real research environment and receive personalized mentorship