@article{FriouxSchaubSchellhornetal.2019,
  author    = {Frioux, Cl{\´e}mence and Schaub, Torsten H. and Schellhorn, Sebastian and Siegel, Anne and Wanko, Philipp},
  title     = {Hybrid metabolic network completion},
  series = {Theory and practice of logic programming},
  volume    = {19},
  journal   = {Theory and practice of logic programming},
  number    = {1},
  publisher = {Cambridge University Press},
  address   = {New York},
  issn      = {1471-0684},
  doi       = {10.1017/S1471068418000455},
  pages     = {83 -- 108},
  year      = {2019},
  abstract  = {Metabolic networks play a crucial role in biology since they capture all chemical reactions in an organism. While there are networks of high quality for many model organisms, networks for less studied organisms are often of poor quality and suffer from incompleteness. To this end, we introduced in previous work an answer set programming (ASP)-based approach to metabolic network completion. Although this qualitative approach allows for restoring moderately degraded networks, it fails to restore highly degraded ones. This is because it ignores quantitative constraints capturing reaction rates. To address this problem, we propose a hybrid approach to metabolic network completion that integrates our qualitative ASP approach with quantitative means for capturing reaction rates. We begin by formally reconciling existing stoichiometric and topological approaches to network completion in a unified formalism. With it, we develop a hybrid ASP encoding and rely upon the theory reasoning capacities of the ASP system dingo for solving the resulting logic program with linear constraints over reals. We empirically evaluate our approach by means of the metabolic network of Escherichia coli. Our analysis shows that our novel approach yields greatly superior results than obtainable from purely qualitative or quantitative approaches.},
  language  = {en}
}
@misc{HalfpapSchlosser2019,
  author    = {Halfpap, Stefan and Schlosser, Rainer},
  title     = {Workload-Driven Fragment Allocation for Partially Replicated Databases Using Linear Programming},
  series = {2019 IEEE 35th International Conference on Data Engineering (ICDE)},
  journal   = {2019 IEEE 35th International Conference on Data Engineering (ICDE)},
  publisher = {IEEE},
  address   = {New York},
  isbn      = {978-1-5386-7474-1},
  issn      = {1084-4627},
  doi       = {10.1109/ICDE.2019.00188},
  pages     = {1746 -- 1749},
  year      = {2019},
  abstract  = {In replication schemes, replica nodes can process read-only queries on snapshots of the master node without violating transactional consistency. By analyzing the workload, we can identify query access patterns and replicate data depending to its access frequency. In this paper, we define a linear programming (LP) model to calculate the set of partial replicas with the lowest overall memory capacity while evenly balancing the query load. Furthermore, we propose a scalable decomposition heuristic to calculate solutions for larger problem sizes. While guaranteeing the same performance as state-of-the-art heuristics, our decomposition approach calculates allocations with up to 23\% lower memory footprint for the TPC-H benchmark.},
  language  = {en}
}
@article{Reich2013,
  author    = {Reich, Sebastian},
  title     = {A nonparametric ensemble transform method for bayesian inference},
  series = {SIAM journal on scientific computing},
  volume    = {35},
  journal   = {SIAM journal on scientific computing},
  number    = {4},
  publisher = {Society for Industrial and Applied Mathematics},
  address   = {Philadelphia},
  issn      = {1064-8275},
  doi       = {10.1137/130907367},
  pages     = {A2013 -- A2024},
  year      = {2013},
  abstract  = {Many applications, such as intermittent data assimilation, lead to a recursive application of Bayesian inference within a Monte Carlo context. Popular data assimilation algorithms include sequential Monte Carlo methods and ensemble Kalman filters (EnKFs). These methods differ in the way Bayesian inference is implemented. Sequential Monte Carlo methods rely on importance sampling combined with a resampling step, while EnKFs utilize a linear transformation of Monte Carlo samples based on the classic Kalman filter. While EnKFs have proven to be quite robust even for small ensemble sizes, they are not consistent since their derivation relies on a linear regression ansatz. In this paper, we propose another transform method, which does not rely on any a priori assumptions on the underlying prior and posterior distributions. The new method is based on solving an optimal transportation problem for discrete random variables.},
  language  = {en}
}