-
Optimal Real-time Bidding Policies for Contract Fulfillment in Second Price Auctions
Authors:
R. J. Kinnear,
R. R. Mazumdar,
P. Marbach
Abstract:
We study a real-time bidding problem resulting from a set of contractual obligations stipulating that a firm win a specified number of heterogeneous impressions or ad placements over a defined duration in a real-time auction. The contracts specify item targeting criteria (which may be overlapping), and a supply requirement. Using the Pontryagin maximum principle, we show that the resulting continu…
▽ More
We study a real-time bidding problem resulting from a set of contractual obligations stipulating that a firm win a specified number of heterogeneous impressions or ad placements over a defined duration in a real-time auction. The contracts specify item targeting criteria (which may be overlapping), and a supply requirement. Using the Pontryagin maximum principle, we show that the resulting continuous time and time inhomogeneous planning problem can be reduced into a finite dimensional convex optimization problem and solved to optimality. In addition, we provide algorithms to update the bidding plan over time via a receding horizon. Finally, we provide numerical results based on real data and show a connection to production-transportation problems.
△ Less
Submitted 17 December, 2020;
originally announced December 2020.
-
Graph Topological Aspects of Granger Causal Network Learning
Authors:
R. J. Kinnear,
R. R. Mazumdar
Abstract:
We study Granger causality in the context of wide-sense stationary time series, where our focus is on the topological aspects of the underlying causality graph. We establish sufficient conditions (in particular, we develop the notion of a "strongly causal" graph topology) under which the true causality graph can be recovered via pairwise causality testing alone, and provide examples from the gene…
▽ More
We study Granger causality in the context of wide-sense stationary time series, where our focus is on the topological aspects of the underlying causality graph. We establish sufficient conditions (in particular, we develop the notion of a "strongly causal" graph topology) under which the true causality graph can be recovered via pairwise causality testing alone, and provide examples from the gene regulatory network literature suggesting that our concept of a strongly causal graph may be applicable to this field. We implement and detail finite-sample heuristics derived from our theory, and establish through simulation the efficiency gains (both statistical and computational) which can be obtained (in comparison to LASSO-type algorithms) when structural assumptions are met.
△ Less
Submitted 16 November, 2019;
originally announced November 2019.
-
Randomized Assignment of Jobs to Servers in Heterogeneous Clusters of Shared Servers for Low Delay
Authors:
Arpan Mukhopadhyay,
A. Karthik,
Ravi R. Mazumdar
Abstract:
We consider the job assignment problem in a multi-server system consisting of $N$ parallel processor sharing servers, categorized into $M$ ($\ll N$) different types according to their processing capacity or speed. Jobs of random sizes arrive at the system according to a Poisson process with rate $N λ$. Upon each arrival, a small number of servers from each type is sampled uniformly at random. The…
▽ More
We consider the job assignment problem in a multi-server system consisting of $N$ parallel processor sharing servers, categorized into $M$ ($\ll N$) different types according to their processing capacity or speed. Jobs of random sizes arrive at the system according to a Poisson process with rate $N λ$. Upon each arrival, a small number of servers from each type is sampled uniformly at random. The job is then assigned to one of the sampled servers based on a selection rule. We propose two schemes, each corresponding to a specific selection rule that aims at reducing the mean sojourn time of jobs in the system.
We first show that both methods achieve the maximal stability region. We then analyze the system operating under the proposed schemes as $N \to \infty$ which corresponds to the mean field. Our results show that asymptotic independence among servers holds even when $M$ is finite and exchangeability holds only within servers of the same type. We further establish the existence and uniqueness of stationary solution of the mean field and show that the tail distribution of server occupancy decays doubly exponentially for each server type. When the estimates of arrival rates are not available, the proposed schemes offer simpler alternatives to achieving lower mean sojourn time of jobs, as shown by our numerical studies.
△ Less
Submitted 20 February, 2015;
originally announced February 2015.
