Showing 1–2 of 2 results for author: Waldherr, S
-
Numerical Gaussian process Kalman filtering
Authors:
Armin Küper,
Steffen Waldherr
Abstract:
In this manuscript we introduce numerical Gaussian process Kalman filtering (GPKF). Numerical Gaussian processes have recently been developed to simulate spatiotemporal models. The contribution of this paper is to embed numerical Gaussian processes into the recursive Kalman filter equations. This embedding enables us to do Kalman filtering on infinite-dimensional systems using Gaussian processes.…
▽ More
In this manuscript we introduce numerical Gaussian process Kalman filtering (GPKF). Numerical Gaussian processes have recently been developed to simulate spatiotemporal models. The contribution of this paper is to embed numerical Gaussian processes into the recursive Kalman filter equations. This embedding enables us to do Kalman filtering on infinite-dimensional systems using Gaussian processes. This is possible because i) we are obtaining a linear model from numerical Gaussian processes, and ii) the states of this model are by definition Gaussian distributed random variables. Convenient properties of the numerical GPKF are that no spatial discretization of the model is necessary, and manual setting up of the Kalman filter, that is fine-tuning the process and measurement noise levels by hand is not required, as they are learned online from the data stream. We showcase the capability of the numerical GPKF in a simulation study of the advection equation.
△ Less
Submitted 11 May, 2020; v1 submitted 3 December, 2019;
originally announced December 2019.
-
Competitive Equilibria in Combinatorial Exchanges with Financially Constrained Buyers:Computational Hardness and Algorithmic Solutions
Authors:
Martin Bichler,
Stefan Waldherr
Abstract:
Advances in computational optimization allow for the organization of large combinatorial markets. We aim for allocations and competitive equilibrium prices, i.e. outcomes that are in the core. The research is motivated by the design of environmental markets, but similar problems appear in energy and logistics markets or in the allocation of airport time slots. Budget constraints are an important c…
▽ More
Advances in computational optimization allow for the organization of large combinatorial markets. We aim for allocations and competitive equilibrium prices, i.e. outcomes that are in the core. The research is motivated by the design of environmental markets, but similar problems appear in energy and logistics markets or in the allocation of airport time slots. Budget constraints are an important concern in many of these markets. While the allocation problem in combinatorial exchanges is already NP-hard with payoff- maximizing bidders, we find that the allocation and pricing problem becomes even $Σ_2^p$-hard if buyers are financially constrained. We introduce mixed integer bilevel linear programs (MIBLP) to compute core prices, and propose pricing functions based on the least core if the core is empty. We also discuss restricted but simpler cases and effective computational techniques for the problem. In numerical experiments we show that in spite of the computational hardness of these problems, we can hope to solve practical problem sizes, in particular if we restrict the size of the coalitions considered in the core computations.
△ Less
Submitted 22 July, 2018;
originally announced July 2018.
