eduzhai > Applied Sciences > Computer Science >

Combining Outcome-Based and Preference-Based Matching A Constrained Priority Mechanism

  • KanKan
  • (0) Download
  • 20210425
  • Save

... pages left unread,continue reading

Document pages: 57 pages

Abstract: We introduce a constrained priority mechanism that combines outcome-basedmatching from machine-learning with preference-based allocation schemes commonin market design. Using real-world data, we illustrate how our mechanism couldbe applied to the assignment of refugee families to host country locations, andkindergarteners to schools. Our mechanism allows a planner to first specify athreshold $ bar g$ for the minimum acceptable average outcome score that shouldbe achieved by the assignment. In the refugee matching context, this scorecorresponds to the predicted probability of employment, while in the studentassignment context it corresponds to standardized test scores. The mechanism isa priority mechanism that considers both outcomes and preferences by assigningagents (refugee families, students) based on their preferences, but subject tomeeting the planner s specified threshold. The mechanism is both strategy-proofand constrained efficient in that it always generates a matching that is notPareto dominated by any other matching that respects the planner s threshold.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×