Decision preserving sorting finds variables that matter in constrained tasks
DePICT: Decision-Preserving Interface for Constrained Downstream Tasks
Machine Learning
Summary
When a computer makes decisions based on many inputs, sometimes changing some inputs a little does not change the final choice. The authors introduce DePICT, a method to find which inputs actually affect the outcome and which do not. They use math that looks at how sensitive the decision is to each input and remove those that do not change the final decision. Their tests show DePICT can find the key inputs exactly and greatly improve decision accuracy compared to other methods.
What this means in practice
- •For operations research teams: Identify essential input factors in complex constrained optimization models to simplify decision interfaces and reduce computational burden.
- •For automated diagnostics engineers: Develop tools that pinpoint which sensor readings actually impact diagnosis decisions, improving efficiency and interpretability.
Authors
Utkarsh Grover, Ravi Ranjan, Agoritsa Polyzou, Wyatt T. Mackey, J. Morris Chang, Leonardo Bobadilla, Xiaomin Lin
Abstract
A constrained optimization problem may involve a parameter in its objective and active constraints, yet the final decision may remain insensitive to small changes in that parameter. This raises a fundamental question: which inputs does a decision making system truly depend on? Building on this question, we introduce DePICT, a procedure for constructing decision preserving interfaces by ranking context directions according to the optimizer's solution sensitivity and aggregating them across an operating regime. We study this problem in a high dimensional setting where primitive context parameterizes a constrained task and the downstream agent observes only a selected subset of context directions. For locally regular constrained programs, we derive a Karush Kuhn Tucker (KKT) based characterization of when a context direction is optimizer relevant. Our analysis shows that appearing in the active optimization problem does not necessarily imply that a variable affects the final decision. Some context directions can alter the KKT conditions while leaving the optimal solution unchanged because their effect is absorbed by the dual variables. DePICT is designed to remove exactly these directions. In a controlled diagnosis, it recovers the decision relevant interface exactly and reduces linear predictor regret to 0.009, compared with 0.475 for the strongest competing baseline.