Source author record

Yuko Sakurai

Yuko Sakurai appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Computer Science and Game Theory

Catalog footprint

What is connected

3works

1topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2020arXiv

Gini Index based Initial Coin Offering Mechanism

As a fundraising method, Initial Coin Offering (ICO) has raised billions of dollars for thousands of startups in the past two years. Existing ICO mechanisms place more emphasis on the short-term benefits of maximal fundraising while ignoring the problem of unbalanced token allocation, which negatively impacts subsequent fundraising and has bad effects on introducing new investors and resources. We propose a new ICO mechanism which uses the concept of Gini index for the very first time as a mechanism design constraint to control allocation inequality. Our mechanism maintains an elegant and straightforward structure. It allows the agents to modify their bids as a price discovery process, while limiting the bids of whales. We analyze the agents' equilibrium behaviors under our mechanism. Under natural technical assumptions, we show that most agents have simple dominant strategies and the equilibrium revenue approaches the optimal revenue asymptotically in the number of agents. We verify our mechanism using real ICO dataset we collected, and confirm that our mechanism performs well in terms of both allocation fairness and revenue.

preprint2020arXiv

Mechanism Design for Public Projects via Neural Networks

We study mechanism design for nonexcludable and excludable binary public project problems. We aim to maximize the expected number of consumers and the expected social welfare. For the nonexcludable public project model, we identify a sufficient condition on the prior distribution for the conservative equal costs mechanism to be the optimal strategy-proof and individually rational mechanism. For general distributions, we propose a dynamic program that solves for the optimal mechanism. For the excludable public project model, we identify a similar sufficient condition for the serial cost sharing mechanism to be optimal for $2$ and $3$ agents. We derive a numerical upper bound. Experiments show that for several common distributions, the serial cost sharing mechanism is close to optimality. The serial cost sharing mechanism is not optimal in general. We design better performing mechanisms via neural networks. Our approach involves several technical innovations that can be applied to mechanism design in general. We interpret the mechanisms as price-oriented rationing-free (PORF) mechanisms, which enables us to move the mechanism's complex (e.g., iterative) decision making off the network, to a separate program. We feed the prior distribution's analytical form into the cost function to provide quality gradients for training. We use supervision to manual mechanisms as a systematic way for initialization. Our approach of "supervision and then gradient descent" is effective for improving manual mechanisms' performances. It is also effective for fixing constraint violations for heuristic-based mechanisms that are infeasible.

preprint2013arXiv

A Dynamic Programming Model for Determining Bidding Strategies in Sequential Auctions: Quasi-linear Utility and Budget Constraints

In this paper, we develop a new method for finding an optimal biddingstrategy in sequential auctions, using a dynamic programming technique. Theexisting method assumes that the utility of a user is represented in anadditive form. Thus, the remaining endowment of money must be explicitlyrepresented in each state, and the calculation of the optimal biddingstrategy becomes time-consuming when the initial endowment of money mbecomes large.In this paper, we develop a new problem formalization that avoids explicitlyrepresenting the remaining endowment, by assuming the utility of a user canbe represented in a quasi-linear form, and representing the payment as astate-transition cost. Experimental evaluations show that we can obtainmore than an m-fold speed-up in the computation time. Furthermore, we havedeveloped a method for obtaining a semi-optimal bidding strategy underbudget constraints, and have experimentally confirmed the efficacy of thismethod.