DeepAI AI Chat
Log In Sign Up

Online Geometric Discrepancy for Stochastic Arrivals with Applications to Envy Minimization

by   Haotian Jiang, et al.

Consider a unit interval [0,1] in which n points arrive one-by-one independently and uniformly at random. On arrival of a point, the problem is to immediately and irrevocably color it in {+1,-1} while ensuring that every interval [a,b] ⊆ [0,1] is nearly-balanced. We define discrepancy as the largest imbalance of any interval during the entire process. If all the arriving points were known upfront then we can color them alternately to achieve a discrepancy of 1. What is the minimum possible expected discrepancy when we color the points online? We show that the discrepancy of the above problem is sub-polynomial in n and that no algorithm can achieve a constant discrepancy. This is a substantial improvement over the trivial random coloring that only gets an O(√(n)) discrepancy. We then obtain similar results for a natural generalization of this problem to 2-dimensions where the points arrive uniformly at random in a unit square. This generalization allows us to improve recent results of Benade et al.<cit.> for the online envy minimization problem when the arrivals are stochastic.


page 1

page 2

page 3

page 4


Online Vector Balancing and Geometric Discrepancy

We consider an online vector balancing question where T vectors, chosen ...

A Sharp Discrepancy Bound for Jittered Sampling

For m, d ∈ℕ, a jittered sampling point set P having N = m^d points in [0...

Online Discrepancy Minimization via Persistent Self-Balancing Walks

We study the online discrepancy minimization problem for vectors in ℝ^d ...

A Unified Approach to Discrepancy Minimization

We study a unified approach and algorithm for constructive discrepancy m...

Online Carpooling using Expander Decompositions

We consider the online carpooling problem: given n vertices, a sequence ...

A practical algorithm to calculate Cap Discrepancy

Uniform distribution of the points has been of interest to researchers f...

Geometric Barriers for Stable and Online Algorithms for Discrepancy Minimization

For many computational problems involving randomness, intricate geometri...