Search

We study the metric facility location problem with client insertions and deletions. This setting differs from the classic dynamic facility location problem, where the set of clients remains the same, but the metric space can change over time. We show a deterministic algorithm that maintains a constant factor approximation to the optimal solution in worst-case time $\tilde O(2^{O(\kappa^2)})$ per client insertion or deletion in metric spaces while answering queries about the cost in $O(1)$ time, where $\kappa$ denotes the doubling dimension of the metric. For metric spaces with bounded doubling dimension, the update time is polylogarithmic in the parameters of the problem., Comment: An extended abstract appeared at the 26th Annual European Symposium on Algorithms (ESA) 2018

Clustering is one of the most fundamental problems in unsupervised learning with a large number of applications. However, classical clustering algorithms assume that the data is static, thus failing to capture many real-world applications where data is constantly changing and evolving. Driven by this, we study the metric $k$-center clustering problem in the fully dynamic setting, where the goal is to efficiently maintain a clustering while supporting an intermixed sequence of insertions and deletions of points. This model also supports queries of the form (1) report whether a given point is a center or (2) determine the cluster a point is assigned to. We present a deterministic dynamic algorithm for the $k$-center clustering problem that provably achieves a $(2+\epsilon)$-approximation in poly-logarithmic update and query time, if the underlying metric has bounded doubling dimension, its aspect ratio is bounded by a polynomial and $\epsilon$ is a constant. An important feature of our algorithm is that the update and query times are independent of $k$. We confirm the practical relevance of this feature via an extensive experimental study which shows that for values of $k$ and $\epsilon$ suggested by theory, our algorithmic construction outperforms the state-of-the-art algorithm in terms of solution quality and running time., Comment: This version includes an experimental evaluation of our algorithm

In this paper, we study the problem of opening centers to cluster a set of clients in a metric space so as to minimize the sum of the costs of the centers and of the cluster radii, in a dynamic environment where clients arrive and depart, and the solution must be updated efficiently while remaining competitive with respect to the current optimal solution. We call this dynamic sum-of-radii clustering problem. We present a data structure that maintains a solution whose cost is within a constant factor of the cost of an optimal solution in metric spaces with bounded doubling dimension and whose worst-case update time is logarithmic in the parameters of the problem., Comment: 10 pages, ESA 2017

The shortest augmenting path technique is one of the fundamental ideas used in maximum matching and maximum flow algorithms. Since being introduced by Edmonds and Karp in 1972, it has been widely applied in many different settings. Surprisingly, despite this extensive usage, it is still not well understood even in the simplest case: online bipartite matching problem on trees. In this problem a bipartite tree $T=(W \uplus B, E)$ is being revealed online, i.e., in each round one vertex from $B$ with its incident edges arrives. It was conjectured by Chaudhuri et. al. [K. Chaudhuri, C. Daskalakis, R. D. Kleinberg, and H. Lin. Online bipartite perfect matching with augmentations. In INFOCOM 2009] that the total length of all shortest augmenting paths found is $O(n \log n)$. In this paper, we prove a tight $O(n \log n)$ upper bound for the total length of shortest augmenting paths for trees improving over $O(n \log^2 n)$ bound [B. Bosek, D. Leniowski, P. Sankowski, and A. Zych. Shortest augmenting paths for online matchings on trees. In WAOA 2015]., Comment: 22 pages, 10 figures