<?xml version="1.0" encoding="utf-8"?>
<export-example>
  <doc>
    <id>7802</id>
    <completedYear/>
    <publishedYear/>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>reportzib</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>2020-03-30</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">The SCIP Optimization Suite 7.0</title>
    <abstract language="eng">The SCIP Optimization Suite provides a collection of software packages for&#13;
mathematical optimization centered around the constraint integer programming frame-&#13;
work SCIP. This paper discusses enhancements and extensions contained in version 7.0&#13;
of the SCIP Optimization Suite. The new version features the parallel presolving library&#13;
PaPILO as a new addition to the suite. PaPILO 1.0 simplifies mixed-integer linear op-&#13;
timization problems and can be used stand-alone or integrated into SCIP via a presolver&#13;
plugin. SCIP 7.0 provides additional support for decomposition algorithms. Besides im-&#13;
provements in the Benders’ decomposition solver of SCIP, user-defined decomposition&#13;
structures can be read, which are used by the automated Benders’ decomposition solver&#13;
and two primal heuristics. Additionally, SCIP 7.0 comes with a tree size estimation&#13;
that is used to predict the completion of the overall solving process and potentially&#13;
trigger restarts. Moreover, substantial performance improvements of the MIP core were&#13;
achieved by new developments in presolving, primal heuristics, branching rules, conflict&#13;
analysis, and symmetry handling. Last, not least, the report presents updates to other&#13;
components and extensions of the SCIP Optimization Suite, in particular, the LP solver&#13;
SoPlex and the mixed-integer semidefinite programming solver SCIP-SDP.</abstract>
    <identifier type="issn">1438-0064</identifier>
    <identifier type="urn">urn:nbn:de:0297-zib-78023</identifier>
    <author>Gerald Gamrath</author>
    <submitter>Felipe Serrano</submitter>
    <author>Daniel Anderson</author>
    <author>Ksenia Bestuzheva</author>
    <author>Wei-Kun Chen</author>
    <author>Leon Eifler</author>
    <author>Maxime Gasse</author>
    <author>Patrick Gemander</author>
    <author>Ambros Gleixner</author>
    <author>Leona Gottwald</author>
    <author>Katrin Halbig</author>
    <author>Gregor Hendel</author>
    <author>Christopher Hojny</author>
    <author>Thorsten Koch</author>
    <author>Pierre Le Bodic</author>
    <author>Stephen J. Maher</author>
    <author>Frederic Matter</author>
    <author>Matthias Miltenberger</author>
    <author>Erik Mühmer</author>
    <author>Benjamin Müller</author>
    <author>Marc Pfetsch</author>
    <author>Franziska Schlösser</author>
    <author>Felipe Serrano</author>
    <author>Yuji Shinano</author>
    <author>Christine Tawfik</author>
    <author>Stefan Vigerske</author>
    <author>Fabian Wegscheider</author>
    <author>Dieter Weninger</author>
    <author>Jakob Witzig</author>
    <series>
      <title>ZIB-Report</title>
      <number>20-10</number>
    </series>
    <collection role="msc" number="65-XX">NUMERICAL ANALYSIS</collection>
    <collection role="msc" number="90-XX">OPERATIONS RESEARCH, MATHEMATICAL PROGRAMMING</collection>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="institutes" number="mip">Mathematical Optimization Methods</collection>
    <collection role="persons" number="gamrath">Gamrath, Gerald</collection>
    <collection role="persons" number="robert.gottwald">Gottwald, Robert</collection>
    <collection role="persons" number="hendel">Hendel, Gregor</collection>
    <collection role="persons" number="koch">Koch, Thorsten</collection>
    <collection role="persons" number="miltenberger">Miltenberger, Matthias</collection>
    <collection role="persons" number="shinano">Shinano, Yuji</collection>
    <collection role="persons" number="vigerske">Vigerske, Stefan</collection>
    <collection role="projects" number="ASTfSCM">ASTfSCM</collection>
    <collection role="projects" number="MIP-ZIBOPT">MIP-ZIBOPT</collection>
    <collection role="projects" number="MODAL-SynLab">MODAL-SynLab</collection>
    <collection role="projects" number="Siemens">Siemens</collection>
    <collection role="projects" number="MODAL-Gesamt">MODAL-Gesamt</collection>
    <collection role="persons" number="schloesser">Schlösser, Franziska</collection>
    <collection role="persons" number="bestuzheva">Bestuzheva, Ksenia</collection>
    <collection role="projects" number="plan4res">Plan4res</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
    <file>https://opus4.kobv.de/opus4-zib/files/7802/scipopt-70.pdf</file>
  </doc>
  <doc>
    <id>7814</id>
    <completedYear/>
    <publishedYear/>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>reportzib</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>2020-04-03</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Estimating the Size of Branch-And-Bound Trees</title>
    <abstract language="eng">This paper investigates the estimation of the size of Branch-and-Bound (B&amp;B) trees for solving mixed-integer programs. We first prove that the size of the B&amp;B tree cannot be approximated within a factor of~2 for general binary programs, unless P equals NP. Second, we review measures of the progress of the B&amp;B search, such as the gap, and propose a new measure, which we call leaf frequency.&#13;
&#13;
We study two simple ways to transform these progress measures into B&amp;B tree size estimates, either as a direct projection, or via double-exponential smoothing, a standard time-series forecasting technique. We then combine different progress measures and their trends into nontrivial estimates using Machine Learning techniques, which yields more precise estimates than any individual measure. The best method we have identified uses all individual measures as features of a random forest model.&#13;
In a large computational study, we train and validate all methods on the publicly available MIPLIB and Coral general purpose benchmark sets. On average, the best method estimates B&amp;B tree sizes within a factor of 3 on the set of unseen test instances even during the early stage of the search, and improves in accuracy as the search progresses. It also achieves a factor 2 over the entire search on each out of six additional sets of homogeneous instances we have tested. All techniques are available in version 7 of the branch-and-cut framework SCIP.</abstract>
    <identifier type="issn">1438-0064</identifier>
    <identifier type="urn">urn:nbn:de:0297-zib-78144</identifier>
    <author>Gregor Hendel</author>
    <submitter>Gregor Hendel</submitter>
    <author>Daniel Anderson</author>
    <author>Pierre Le Bodic</author>
    <author>Marc Pfetsch</author>
    <series>
      <title>ZIB-Report</title>
      <number>20-02</number>
    </series>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>mixed integer programming</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>machine learning</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>branch and bound</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>forecasting</value>
    </subject>
    <collection role="msc" number="90-XX">OPERATIONS RESEARCH, MATHEMATICAL PROGRAMMING</collection>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="persons" number="hendel">Hendel, Gregor</collection>
    <collection role="projects" number="ASTfSCM">ASTfSCM</collection>
    <collection role="projects" number="MODAL-SynLab">MODAL-SynLab</collection>
    <collection role="projects" number="Siemens">Siemens</collection>
    <collection role="projects" number="MODAL-Gesamt">MODAL-Gesamt</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
    <file>https://opus4.kobv.de/opus4-zib/files/7814/zibreportestimatingSearchTreeSize.pdf</file>
  </doc>
  <doc>
    <id>7265</id>
    <completedYear/>
    <publishedYear/>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>reportzib</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>2019-02-26</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Clairvoyant Restarts in Branch-and-Bound Search Using Online Tree-Size Estimation</title>
    <abstract language="eng">We propose a simple and general online method to measure the search progress within the Branch-and-Bound algorithm, from which we estimate the size of the remaining search tree. We then show how this information can help solvers algorithmically at runtime by designing a restart strategy for Mixed-Integer Programming (MIP) solvers that decides whether to restart the search based on the current estimate of the number of remaining nodes in the tree. We refer to this type of algorithm as clairvoyant.&#13;
Our clairvoyant restart strategy outperforms a state-of-the-art solver on a large set of publicly available MIP benchmark instances.&#13;
It is implemented in the MIP solver SCIP and will be available in future releases.</abstract>
    <identifier type="issn">1438-0064</identifier>
    <identifier type="urn">urn:nbn:de:0297-zib-72653</identifier>
    <enrichment key="opus.doi.autoCreate">false</enrichment>
    <enrichment key="opus.urn.autoCreate">true</enrichment>
    <author>Daniel Anderson</author>
    <submitter>Gregor Hendel</submitter>
    <author>Gregor Hendel</author>
    <author>Pierre Le Bodic</author>
    <author>Jan Merlin Viernickel</author>
    <series>
      <title>ZIB-Report</title>
      <number>19-11</number>
    </series>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>Mixed-Integer Programming solvers</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>Restart</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>Progress measures</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>tree-size estimates</value>
    </subject>
    <collection role="msc" number="90-XX">OPERATIONS RESEARCH, MATHEMATICAL PROGRAMMING</collection>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="institutes" number="mip">Mathematical Optimization Methods</collection>
    <collection role="persons" number="hendel">Hendel, Gregor</collection>
    <collection role="projects" number="ASTfSCM">ASTfSCM</collection>
    <collection role="projects" number="MIP-ZIBOPT">MIP-ZIBOPT</collection>
    <collection role="projects" number="MODAL-SynLab">MODAL-SynLab</collection>
    <collection role="projects" number="Siemens">Siemens</collection>
    <collection role="projects" number="MODAL-Gesamt">MODAL-Gesamt</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
    <file>https://opus4.kobv.de/opus4-zib/files/7265/zib-report.pdf</file>
  </doc>
  <doc>
    <id>6629</id>
    <completedYear/>
    <publishedYear/>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>reportzib</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>2017-12-21</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">The SCIP Optimization Suite 5.0</title>
    <abstract language="eng">This article describes new features and enhanced algorithms made available in version 5.0 of the SCIP Optimization Suite. In its central component, the constraint integer programming solver SCIP, remarkable performance improvements have been achieved for solving mixed-integer linear and nonlinear programs. On MIPs, SCIP 5.0 is about 41 % faster than SCIP 4.0 and over twice as fast on instances that take at least 100 seconds to solve. For MINLP, SCIP 5.0 is about 17 % faster overall and 23 % faster on instances that take at least 100 seconds to solve. This boost is due to algorithmic advances in several parts of the solver such as cutting plane generation and management, a new adaptive coordination of large neighborhood search heuristics, symmetry handling, and strengthened McCormick relaxations for bilinear terms in MINLPs. Besides discussing the theoretical background and the implementational aspects of these developments, the report describes recent additions for the other software packages connected to SCIP, in particular for the LP solver SoPlex, the Steiner tree solver SCIP-Jack, the MISDP solver SCIP-SDP, and the parallelization framework UG.</abstract>
    <identifier type="issn">1438-0064</identifier>
    <identifier type="urn">urn:nbn:de:0297-zib-66297</identifier>
    <author>Ambros Gleixner</author>
    <submitter>Gregor Hendel</submitter>
    <author>Leon Eifler</author>
    <author>Tristan Gally</author>
    <author>Gerald Gamrath</author>
    <author>Patrick Gemander</author>
    <author>Robert Lion Gottwald</author>
    <author>Gregor Hendel</author>
    <author>Christopher Hojny</author>
    <author>Thorsten Koch</author>
    <author>Matthias Miltenberger</author>
    <author>Benjamin Müller</author>
    <author>Marc Pfetsch</author>
    <author>Christian Puchert</author>
    <author>Daniel Rehfeldt</author>
    <author>Franziska Schlösser</author>
    <author>Felipe Serrano</author>
    <author>Yuji Shinano</author>
    <author>Jan Merlin Viernickel</author>
    <author>Stefan Vigerske</author>
    <author>Dieter Weninger</author>
    <author>Jonas T. Witt</author>
    <author>Jakob Witzig</author>
    <series>
      <title>ZIB-Report</title>
      <number>17-61</number>
    </series>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>constraint integer programming</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>linear programming</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>mixed-integer linear programming</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>mixed-integer nonlinear programming</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>optimization solver</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>branch-and-cut</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>branch-and-price</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>column generation framework</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>parallelization</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>mixed-integer semidefinite programming</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>Steiner tree optimization</value>
    </subject>
    <collection role="msc" number="65-XX">NUMERICAL ANALYSIS</collection>
    <collection role="msc" number="90-XX">OPERATIONS RESEARCH, MATHEMATICAL PROGRAMMING</collection>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="institutes" number="mip">Mathematical Optimization Methods</collection>
    <collection role="persons" number="gamrath">Gamrath, Gerald</collection>
    <collection role="persons" number="robert.gottwald">Gottwald, Robert</collection>
    <collection role="persons" number="hendel">Hendel, Gregor</collection>
    <collection role="persons" number="koch">Koch, Thorsten</collection>
    <collection role="persons" number="miltenberger">Miltenberger, Matthias</collection>
    <collection role="persons" number="rehfeldt">Rehfeldt, Daniel</collection>
    <collection role="persons" number="shinano">Shinano, Yuji</collection>
    <collection role="persons" number="vigerske">Vigerske, Stefan</collection>
    <collection role="projects" number="ASTfSCM">ASTfSCM</collection>
    <collection role="projects" number="MIP-ZIBOPT">MIP-ZIBOPT</collection>
    <collection role="projects" number="MODAL-SynLab">MODAL-SynLab</collection>
    <collection role="projects" number="Siemens">Siemens</collection>
    <collection role="projects" number="MODAL-Gesamt">MODAL-Gesamt</collection>
    <collection role="persons" number="schloesser">Schlösser, Franziska</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
    <file>https://opus4.kobv.de/opus4-zib/files/6629/scipopt-50.pdf</file>
  </doc>
  <doc>
    <id>7231</id>
    <completedYear/>
    <publishedYear>2018</publishedYear>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>article</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>--</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Adaptive Large Neighborhood Search for Mixed Integer Programming</title>
    <abstract language="eng">Large Neighborhood Search (LNS) heuristics are among the most powerful but also most expensive heuristics for mixed integer programs (MIP). Ideally, a solver learns adaptively which LNS heuristics work best for the MIP problem at hand in order to concentrate its limited computational budget. To this end, this work introduces Adaptive Large Neighborhood Search (ALNS) for MIP, a primal heuristic that acts a framework for eight popular LNS heuristics such as Local Branching and Relaxation Induced Neighborhood Search (RINS). We distinguish the available LNS heuristics by their individual search domains, which we call neighborhoods. The decision which neighborhood should be executed is guided by selection strategies for the multi armed bandit problem, a related optimization problem during which suitable actions have to be chosen to maximize a reward function. In this paper, we propose an LNS-specific reward function to learn to distinguish between the available neighborhoods based on successful calls and failures. A second, algorithmic enhancement is a generic variable fixing priorization, which ALNS employs to adjust the subproblem complexity as needed. This is particularly useful for some neighborhoods which do not fix variables by themselves. The proposed primal heuristic has been implemented within the MIP solver SCIP. An extensive computational study is conducted to compare different LNS strategies within our ALNS framework on a large set of publicly available MIP instances from the MIPLIB and Coral benchmark sets. The results of this simulation are used to calibrate the parameters of the bandit selection strategies. A second computational experiment shows the computational benefits of the proposed ALNS framework within the MIP solver SCIP.</abstract>
    <parentTitle language="eng">Mathematical Programming Computation</parentTitle>
    <enrichment key="PeerReviewed">yes</enrichment>
    <enrichment key="SubmissionStatus">under review</enrichment>
    <enrichment key="PreprintUrn">urn:nbn:de:0297-zib-71167</enrichment>
    <author>Gregor Hendel</author>
    <submitter>Gregor Hendel</submitter>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="persons" number="hendel">Hendel, Gregor</collection>
    <collection role="projects" number="ASTfSCM">ASTfSCM</collection>
    <collection role="projects" number="MIP-ZIBOPT">MIP-ZIBOPT</collection>
    <collection role="projects" number="MODAL-SynLab">MODAL-SynLab</collection>
    <collection role="projects" number="Siemens">Siemens</collection>
    <collection role="projects" number="MODAL-Gesamt">MODAL-Gesamt</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
  </doc>
  <doc>
    <id>7227</id>
    <completedYear/>
    <publishedYear>2018</publishedYear>
    <thesisYearAccepted/>
    <language>deu</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>conferenceobject</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>--</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="deu">Clairvoyant Restarts in Branch-and-Bound Search Using Online Tree-Size Estimation</title>
    <abstract language="deu">We propose a simple and general online method to measure the search progress within the branch-and-bound algorithm, from which we estimate the size of the remaining search tree. We then show how this information can help solvers algorithmically at runtime by designing a restart strategy for Mixed-Integer Programming (MIP) solvers that decides whether to restart the search based on the current estimate of the number of remaining nodes in the tree. We refer to this type of algorithm as clairvoyant. Our clairvoyant restart strategy outperforms a state-of-the-art solver on a large set of publicly available MIP benchmark instances. It is implemented in the MIP solver SCIP and will be available in future releases.</abstract>
    <parentTitle language="deu">AAAI-19: Proceedings of the Thirty-Third AAAI Conference on Artificial Intelligence</parentTitle>
    <enrichment key="PeerReviewed">yes</enrichment>
    <enrichment key="SubmissionStatus">accepted for publication</enrichment>
    <enrichment key="AcceptedDate">2018-10-31</enrichment>
    <enrichment key="PreprintUrn">urn:nbn:de:0297-zib-72653</enrichment>
    <author>Daniel Anderson</author>
    <submitter>Gregor Hendel</submitter>
    <author>Gregor Hendel</author>
    <author>Pierre Le Bodic</author>
    <author>Jan Merlin Viernickel</author>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="institutes" number="mip">Mathematical Optimization Methods</collection>
    <collection role="persons" number="hendel">Hendel, Gregor</collection>
    <collection role="projects" number="ASTfSCM">ASTfSCM</collection>
    <collection role="projects" number="MIP-ZIBOPT">MIP-ZIBOPT</collection>
    <collection role="projects" number="MODAL-SynLab">MODAL-SynLab</collection>
    <collection role="projects" number="Siemens">Siemens</collection>
    <collection role="projects" number="MODAL-Gesamt">MODAL-Gesamt</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
  </doc>
  <doc>
    <id>7116</id>
    <completedYear/>
    <publishedYear/>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>reportzib</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>2018-12-18</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Adaptive Large Neighborhood Search for Mixed Integer Programming</title>
    <abstract language="eng">Large Neighborhood Search (LNS) heuristics are among the most powerful but also most expensive heuristics for mixed integer programs (MIP). Ideally, a solver learns adaptively which LNS heuristics work best for the MIP problem at hand in order to concentrate its limited computational budget.&#13;
&#13;
To this end, this work introduces Adaptive Large Neighborhood Search (ALNS) for MIP, a primal heuristic that acts a framework for eight popular LNS heuristics such as Local Branching and Relaxation Induced Neighborhood Search (RINS). We distinguish the available LNS heuristics by their individual search domains, which we call neighborhoods. The decision which neighborhood should be executed is guided by selection strategies for the multi armed bandit problem, a related optimization problem during which suitable actions have to be chosen to maximize a reward function. In this paper, we propose an LNS-specific reward function to learn to distinguish between the available neighborhoods based on successful calls and failures. A second, algorithmic enhancement is a generic variable fixing priorization, which ALNS employs to adjust the subproblem complexity as needed. This is particularly useful for some neighborhoods which do not fix variables by themselves. The proposed primal heuristic has been implemented&#13;
within the MIP solver SCIP. An extensive computational study is conducted to compare different LNS strategies within our ALNS framework on a large set of publicly available MIP instances from the MIPLIB and Coral benchmark sets. The results of this simulation are used to calibrate the parameters of the bandit selection strategies. A second computational experiment shows the computational benefits of the proposed ALNS framework within the MIP solver SCIP.</abstract>
    <identifier type="issn">1438-0064</identifier>
    <identifier type="urn">urn:nbn:de:0297-zib-71167</identifier>
    <author>Gregor Hendel</author>
    <submitter>Gregor Hendel</submitter>
    <series>
      <title>ZIB-Report</title>
      <number>18-60</number>
    </series>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>mixed integer programming</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>primal heuristics</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>large neighborhood search</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>multi armed bandit problem</value>
    </subject>
    <collection role="msc" number="90-XX">OPERATIONS RESEARCH, MATHEMATICAL PROGRAMMING</collection>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="institutes" number="mip">Mathematical Optimization Methods</collection>
    <collection role="persons" number="hendel">Hendel, Gregor</collection>
    <collection role="projects" number="ASTfSCM">ASTfSCM</collection>
    <collection role="projects" number="MIP-ZIBOPT">MIP-ZIBOPT</collection>
    <collection role="projects" number="MODAL-SynLab">MODAL-SynLab</collection>
    <collection role="projects" number="Siemens">Siemens</collection>
    <collection role="projects" number="MODAL-Gesamt">MODAL-Gesamt</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
    <file>https://opus4.kobv.de/opus4-zib/files/7116/adaptive-lns-master.pdf</file>
  </doc>
  <doc>
    <id>7804</id>
    <completedYear/>
    <publishedYear>2021</publishedYear>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst>443</pageFirst>
    <pageLast>490</pageLast>
    <pageNumber/>
    <edition/>
    <issue>3</issue>
    <volume>13</volume>
    <type>article</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>--</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">MIPLIB 2017: Data-Driven Compilation of the 6th Mixed-Integer Programming Library</title>
    <abstract language="eng">We report on the selection process leading to the sixth version of the Mixed Integer Programming Library. Selected from an initial pool of over 5,000 instances, the new MIPLIB 2017 collection consists of 1,065 instances. A subset of 240 instances was specially selected for benchmarking solver performance. For the first time, the compilation of these sets was done using a data-driven selection process supported by the solution of a sequence of mixed integer optimization problems, which encoded requirements on diversity and balancedness with respect to instance features and performance data.</abstract>
    <parentTitle language="eng">Mathematical Programming Computation</parentTitle>
    <identifier type="doi">10.1007/s12532-020-00194-3</identifier>
    <enrichment key="PeerReviewed">yes</enrichment>
    <enrichment key="AcceptedDate">2020-09-10</enrichment>
    <author>Ambros Gleixner</author>
    <submitter>Ambros Gleixner</submitter>
    <author>Gregor Hendel</author>
    <author>Gerald Gamrath</author>
    <author>Tobias Achterberg</author>
    <author>Michael Bastubbe</author>
    <author>Timo Berthold</author>
    <author>Philipp M. Christophel</author>
    <author>Kati Jarck</author>
    <author>Thorsten Koch</author>
    <author>Jeff Linderoth</author>
    <author>Marco Lübbecke</author>
    <author>Hans Mittelmann</author>
    <author>Derya Ozyurt</author>
    <author>Ted Ralphs</author>
    <author>Domenico Salvagnin</author>
    <author>Yuji Shinano</author>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="institutes" number="mip">Mathematical Optimization Methods</collection>
    <collection role="persons" number="achterberg">Achterberg, Tobias</collection>
    <collection role="persons" number="berthold">Berthold, Timo</collection>
    <collection role="persons" number="gamrath">Gamrath, Gerald</collection>
    <collection role="persons" number="hendel">Hendel, Gregor</collection>
    <collection role="persons" number="koch">Koch, Thorsten</collection>
    <collection role="persons" number="shinano">Shinano, Yuji</collection>
    <collection role="projects" number="ASTfSCM">ASTfSCM</collection>
    <collection role="projects" number="MIP-ZIBOPT">MIP-ZIBOPT</collection>
    <collection role="projects" number="MODAL-SynLab">MODAL-SynLab</collection>
    <collection role="projects" number="Siemens">Siemens</collection>
    <collection role="projects" number="MODAL-Gesamt">MODAL-Gesamt</collection>
    <collection role="institutes" number="ais2t">AI in Society, Science, and Technology</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
  </doc>
  <doc>
    <id>6836</id>
    <completedYear/>
    <publishedYear/>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>reportzib</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>2018-04-10</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Feature-Based Algorithm Selection for Mixed Integer Programming</title>
    <abstract language="eng">Mixed integer programming is a versatile and valuable optimization tool. However, solving specific problem instances can be computationally demanding even for cutting-edge solvers. Such long running times are often significantly reduced by an appropriate change of the solver's parameters. In this paper we investigate "algorithm selection", the task of choosing among a set of algorithms the ones that are likely to perform best for a particular instance. &#13;
&#13;
In our case, we treat different parameter settings of the MIP solver SCIP as different algorithms to choose from. Two peculiarities of the MIP solving process have our special attention. We address the well-known problem of performance variability by using multiple random seeds. Besides solving time, primal dual integrals are recorded as a second performance measure in order to distinguish solvers that timed out.&#13;
&#13;
We collected feature and performance data for a large set of publicly available MIP instances. The algorithm selection problem is addressed by several popular, feature-based methods, which have been partly extended for our purpose. Finally, an analysis of the feature space and performance results of the selected algorithms are presented.</abstract>
    <identifier type="issn">1438-0064</identifier>
    <identifier type="urn">urn:nbn:de:0297-zib-68362</identifier>
    <author>Alexander Georges</author>
    <submitter>Gregor Hendel</submitter>
    <author>Ambros Gleixner</author>
    <author>Gorana Gojic</author>
    <author>Robert Lion Gottwald</author>
    <author>David Haley</author>
    <author>Gregor Hendel</author>
    <author>Bartlomiej Matejczyk</author>
    <series>
      <title>ZIB-Report</title>
      <number>18-17</number>
    </series>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>algorithm selection</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>mixed integer programming</value>
    </subject>
    <collection role="msc" number="90-XX">OPERATIONS RESEARCH, MATHEMATICAL PROGRAMMING</collection>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="institutes" number="mip">Mathematical Optimization Methods</collection>
    <collection role="persons" number="robert.gottwald">Gottwald, Robert</collection>
    <collection role="persons" number="hendel">Hendel, Gregor</collection>
    <collection role="projects" number="ASTfSCM">ASTfSCM</collection>
    <collection role="projects" number="MODAL-SynLab">MODAL-SynLab</collection>
    <collection role="projects" number="Siemens">Siemens</collection>
    <collection role="projects" number="BEAM-ME">BEAM-ME</collection>
    <collection role="projects" number="MODAL-Gesamt">MODAL-Gesamt</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
    <file>https://opus4.kobv.de/opus4-zib/files/6836/main.pdf</file>
  </doc>
  <doc>
    <id>7113</id>
    <completedYear/>
    <publishedYear>2018</publishedYear>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>conferenceobject</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>--</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Adaptive Algorithmic Behavior for Solving Mixed Integer Programs Using Bandit Algorithms</title>
    <abstract language="eng">State-of-the-art solvers for mixed integer programs (MIP) govern a variety of algorithmic components. Ideally, the solver adaptively learns to concentrate its computational budget on those components that perform well on a particular problem, especially if they are time consuming. We focus on three such algorithms, namely the classes of large neighborhood search and diving heuristics as well as Simplex pricing strategies. For each class we propose a selection strategy that is updated based on the observed runtime behavior, aiming to ultimately select only the best algorithms for a given instance. We review several common strategies for such a selection scenario under uncertainty, also known as Multi Armed Bandit Problem. In order to apply those bandit strategies, we carefully design reward functions to rank and compare each individual heuristic or pricing algorithm within its respective class. Finally, we discuss the computational benefits of using the proposed adaptive selection within the SCIP Optimization Suite on publicly available MIP instances.</abstract>
    <parentTitle language="eng">OR 2018: International Conference on Operations Research</parentTitle>
    <enrichment key="Series">Operations Research 2018 Proceedings</enrichment>
    <enrichment key="PeerReviewed">yes</enrichment>
    <enrichment key="SubmissionStatus">accepted for publication</enrichment>
    <enrichment key="AcceptedDate">2018-11-16</enrichment>
    <enrichment key="PreprintUrn">urn:nbn:de:0297-zib-69563</enrichment>
    <author>Gregor Hendel</author>
    <submitter>Gregor Hendel</submitter>
    <author>Matthias Miltenberger</author>
    <author>Jakob Witzig</author>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="institutes" number="mip">Mathematical Optimization Methods</collection>
    <collection role="persons" number="hendel">Hendel, Gregor</collection>
    <collection role="persons" number="miltenberger">Miltenberger, Matthias</collection>
    <collection role="projects" number="ASTfSCM">ASTfSCM</collection>
    <collection role="projects" number="MIP-ZIBOPT">MIP-ZIBOPT</collection>
    <collection role="projects" number="MODAL-SynLab">MODAL-SynLab</collection>
    <collection role="projects" number="Siemens">Siemens</collection>
    <collection role="projects" number="MODAL-Gesamt">MODAL-Gesamt</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
  </doc>
</export-example>
