Leximine ness reduces the picture for the utilitarian Optim ptimization in their paper, Eden Hartman, Yonaton Uman Neman, Avinatan Hasidim and Arel Segal-Halvi Present a plan for solving social selection problems. In this interview, Eden tells us more about such problems, team methods and why there is such an interesting and challenging area for study.
What is the topic of research in your paper?
The paper sees social selection problems – situations where people’s groups (called agents) must make a decision that affects everyone. For example, imagine that we need to decide how to share inheritance in many heirs. Each agent has its own choices on potential consequences, and the goal is to choose the “best” result for the whole society. But how should we define what “best” is for society? There are many potential definitions.
Two frequent and often contrasting definitions are the best utilitarians, which focus on maximizing total welfare (ie, sum of utilities); And the best of equality, which is focused on maximizing the utility. Laximine best normalizes the equalityist. Its first purpose is to maximize the utility; After that, in all those options that maximize the utility, it chooses one that maximizes the second most important utility-the lowest utility, and so forth.
Although leximine is generally regarded as a better approach than a utilitarian, it falls at the cost of calculation. The calculation of a choice that maximizes the utilitarian welfare is easier than to maximize the welfare of the uniform, while the laximine is usually more complex to find out.
In paper, we introduce a general reduction in the utilitarian from leximine. In particular, for any social selection problem with non-negative utilities, we prove that black-b Box qs are given (preventionist) results that maximize utilitarian welfare, one can get a lottery algorithm that gives a lottery in view of these results. Our decline extends to estimates and randomized solvers. Overall, with a decrease in our hands, Pattimizing Leximine in anticipation is no more difficult than ptimizing izing to utilitarian welfare.
What were your main findings?
One of the main effects of this paper is that it reveals the connection between the worst individuals and overall social welfare concerns. This challenges the common perception that the Fair painting should always come at the expense of efficiency, and instead suggests that in some cases, both can go hand in hand.
Can you tell us about the effects of your research and why there is an interesting field for that study?
I find social choice problems attractive because they are deply tendy origin in real life situations. The need for a ness pain in the group’s decision is something that we experience-from small-to-be situations, to large-scale decisions that affect the whole population, or the formation of the government, from small-to-be situations to divide toys in children. This makes the field especially attractive, in my view, a combination of real-world intervals with the beauty of mathematics .He is modeling and formal definitions and algorithmics-functions that help us find decisions to satisfy the criteria of satisfaction and welfare.
Today, insights in this area are already being used in real-world systems, and even by governments, decision-making processes and legal structures that promote ness pain, efficiency and truth-playing. I find it incredibly exciting!
High-level description of the reduced algorithm. An arrow from Element A to B indicates that the corresponding section reduces the problem b. White ingredients are applied in paper; Gray ingredients represent existing algorithms; Black component is Black-Box for utilitarian welfare.
Can you explain your method?
The main stages of the research began with a comprehensive literary review-understanding existing algorithms to find Laximine-Renowned solutions, exploring various imagination of estimates, and involving many trials and mistakes. We learned very much from working through solid examples, often asking ourselves: What can we learn from this particular case?
We took step by step. Each time, we found a way to make the problem a little easier, until we finally reached a version that could only be solved using the utilitarian welfare solver.
We brought with many field equipment, such as – outward Optim ptimization, linear programming and dialogue. At one point, approximately evidence also depends on the geometric range arguments. In the end it was really satisfying to see how all the pieces came together, however, we did not know in advance what the way would look like.
I think the most important part wasn’t defeat – even though the solution was not clear. Every failed attempt was taught us something, and it was crucial to the end result of early insights.
What are you doing next to this field?
I really enjoy working in finding a good settlement. For example, in this paper, we proposed a new definition for leximine estimates and showed that it is possible to calculate such estimates, even in cases where finding the best solution is known as NP-Hard.
I am currently working on the research projects I am working on and really excited I am in the field of truth. The goal of this field is to create systems where all participants are encouraged to inform their true choices – or, more accurately, there is at least no incentive for lies. In many cases, it is known that no rule can guarantee everyone’s truth. The approach we are looking for is again, the potential settlement: our goal is to create systems where both are safe (the agent never hurts) and the profitable requires a lot of effort – in particular, the manipulator will need to collect detailed information about the choices of others.
More widely, I think my approach to future research is to recognize that not everything is black and white. The fact that we can’t achieve something does not mean that we can’t achieve anything – asking what is in Rich comes closer to us, one step at a time.
About Eden
 | Eden Hartman is a third year’s PhD student at computer science at Bar-Elan University in Israel. She has been advised by Professor Avinatan Hasidim (University of Bar-Elan and Google), Professor Yonnat Uman Man (University of Bar-Elan) and Professor Areal Segal-Helvi (Ariel University). Its research focuses on the social selection of the calculation, especially on the problems of the reasonable department. Its function often includes approximately concepts, exploring meaningful settlement and algorithmic development and analysis. |
Read the full work
Reducing leximine ness pain for utilitarian Optim ptimization, Eden Hartman, Yonaton Uman Mind, Avinatan Hasidim, Arel Segal-HalviAAAI 2025.
TGS GS: AAAI, AAAI 2025
Lucy Smith is a senior managing editor for AIHUB.
