Fair Matching Systems Can Still Produce Unequal Outcomes, New Research Finds
A computerized matching system can be designed to be fair and still produce unequal outcomes if the people using it do not understand how it works.
The INFORMS Analytics+ Conference showcased how companies are turning advanced analytics and AI into measurable operational results
In most emergency departments, beds are a hot commodity. But some patients don’t actually need a bed for their treatment – and accurately identifying those people is a low-lift but effective way to increase efficiency without sacrificing care quality, finds a study recently published in the journal Management Science.
The researchers developed a simple decision tree, informed by mathematical and machine learning models, to help staff decide whether a patient truly needs a bed or can be routed to a “vertical processing pathway” – that is, an area where they can be treated sitting up. Many hospitals already have such an area but don’t necessarily have a systematic way to decide which patients are treated there.
Diversity alone won’t get the votes for funding public education. It takes communities that are more integrated and connected, where residents have a sense of shared stake.
An audio journey of how data and analytics save lives, save money and solve problems.

Jeff Cohen
Chief Strategy Officer
INFORMS
Catonsville, MD
[email protected]
443-757-3565
Explore our resources for multiple topics including:
A computerized matching system can be designed to be fair and still produce unequal outcomes if the people using it do not understand how it works.
We are drowning in numbers, and we no longer understand what they mean. Every day, headlines bombard us with figures meant to inform: billions in spending, trillions in debt, percentages signaling growth or decline. But for most of us, these numbers blur together. They register as “large” or “small,” but almost never as real. And that is more than a math problem. It is a civic one.
Big data, artificial intelligence and advanced pricing algorithms make it easier than ever for companies to fine-tune prices for individual products to closely reflect their unique value and cost. The conventional wisdom is straightforward: better data, better algorithms and sharper segmentation should produce better profits. But new research suggests that the most profitable answer isn’t always more fine-grained pricing across a product line. In fact, it is fewer, better-chosen price points.
INFORMS has awarded Microsoft (NASDAQ: MSFT) its 2026 Franz Edelman Award for Achievement in Advanced Analytics, Operations Research and Management Science, for reengineering how global cloud infrastructure is planned and delivered, applying advanced analytics and AI to orchestrate complex fulfillment decisions across its rapidly expanding data center network.
INFORMS has awarded Microsoft, a global leader in cloud computing and technology solutions, the 2026 INFORMS Prize. This prestigious award recognizes Microsoft’s use of advanced analytics and O.R. to optimize its end-to-end cloud infrastructure and capacity management to deliver best-in-class efficiency, scalability and reliability.

OR/MS Today is the INFORMS member magazine that shares the latest research and best practices in operations research, analytics and the management sciences.
Access OR/MS Today Magazine
Analytics magazine showcases articles and research reports based on big data, AI, machine learning, data analytics and other new-age technologies.
Access Analytics Magazine