When:
Monday - December 16, 2024
15:00 PM - 16:00 PM CET
09:00 AM - 10:00 AM EST
19:30 PM - 20:30 PM IST
Speaker:
Marcel Hunting, Algorithmic Capabilities, AIMMS
Introduction:
In optimization, even the most meticulously designed models can become infeasible - not because of errors in formulation, but due to incorrect or unrealistic input data provided by end-users. This can lead to frustration and confusion when users ask, “Why doesn’t this work?” and expect some explanation as well as actionable insights to correct the problem.
This webinar introduces a powerful new feature in AIMMS which tackles the challenge of diagnosing and resolving infeasibility caused by incorrect input data. The focus is on empowering end-users to identify and correct problematic data in a systematic, user-friendly manner, ensuring their optimization models achieve feasible solutions efficiently.
Through this new AIMMS function, users can:
- Detect the root causes of infeasibility: By calculating Irreducible Inconsistent Sets (IIS) or solving feasibility problems, the function pinpoints conflicting constraints.
- Focus on actionable constraints: Non-changeable parameters are filtered out, allowing the user to concentrate on inputs that can realistically be adjusted.
- Leverage “reverse generation” insights: The function identifies which changeable parameters influence infeasible constraints, providing clear guidance.
The process doesn’t stop there. Using intuitive graphical tools within AIMMS:
- Automated suspicion levels (High, Medium, Low): Parameters likely contributing to infeasibility are flagged with suspicion levels.
- Visual cues for user correction: Suspicious data values are highlighted in tables, with color-coded annotations (e.g., ranging from light pink to red) guiding users to problematic inputs.
- Iterative resolution: Users can adjust flagged values, re-solve the model, and refine inputs until feasibility is restored.
This session will demonstrate how to use this AIMMS functionality effectively, showing how the output message and graphical annotations work seamlessly to deliver actionable insights. By the end of the webinar, you’ll be equipped to:
- Analyze infeasibility causes with confidence,
- Leverage AIMMS tools to highlight and correct input data errors,
- Achieve feasible optimization outcomes efficiently, even when starting from challenging data scenarios.
Whether you are an optimization practitioner, a developer of decision-support tools, or an end-user seeking clarity, this webinar will equip you with valuable knowledge to navigate and resolve infeasibility data issues effectively.
Join us to see how this new AIMMS function enhances decision-making by turning infeasibility from a challenge into an opportunity for refining input data quality and improving model outcomes. You will discover how AIMMS makes dealing with infeasibility intuitive, iterative, and impactful - turning complexity into clarity for optimization users.
Registration:
Interested? Please write an e-mail to: Marcel.Hunting@aimms.com. Then we will add you to the webinar mailing list and you will receive a link a few days before the webinar.