lectsched-1: Instance-to-Instance Comparison Results

Type: Instance
Submitter: Harald Schilly
Description: University lecture scheduling instance
MIPLIB Entry

Parent Instance (lectsched-1)

All other instances below were be compared against this "query" instance.

lectsched-1 Raw lectsched-1 Decomposed lectsched-1 Composite of MIC top 5 lectsched-1 Composite of MIPLIB top 5 lectsched-1 Model Group Composite
Raw This is the CCM image before the decomposition procedure has been applied.
Decomposed This is the CCM image after a decomposition procedure has been applied. This is the image used by the MIC's image-based comparisons for this query instance.
Composite of MIC Top 5 Composite of the five decomposed CCM images from the MIC Top 5.
Composite of MIPLIB Top 5 Composite of the five decomposed CCM images from the MIPLIB Top 5.
Model Group Composite Image Composite of the decomposed CCM images for every instance in the same model group as this query.

MIC Top 5 Instances

These are the 5 decomposed CCM images that are most similar to decomposed CCM image for the the query instance, according to the ISS metric.

Decomposed These decomposed images were created by GCG.
lectsched-3 decomposed lectsched-2 decomposed lectsched-5-obj decomposed probportfolio decomposed tanglegram6 decomposed
Name lectsched-3 [MIPLIB] lectsched-2 [MIPLIB] lectsched-5-obj [MIPLIB] probportfolio [MIPLIB] tanglegram6 [MIPLIB]
Rank / ISS The image-based structural similarity (ISS) metric measures the Euclidean distance between the image-based feature vectors for the query instance and all other instances. A smaller ISS value indicates greater similarity.
1 / 0.256 2 / 0.362 3 / 0.364 4 / 0.682 5 / 0.719
Raw These images represent the CCM images in their raw forms (before any decomposition was applied) for the MIC top 5.
lectsched-3 raw lectsched-2 raw lectsched-5-obj raw probportfolio raw tanglegram6 raw

MIPLIB Top 5 Instances

These are the 5 instances that are most closely related to the query instance, according to the instance statistic-based similarity measure employed by MIPLIB 2017

Decomposed These decomposed images were created by GCG.
lectsched-3 decomposed lectsched-2 decomposed neos-3355323-arnon decomposed neos-3211096-shag decomposed neos-3603137-hoteo decomposed
Name lectsched-3 [MIPLIB] lectsched-2 [MIPLIB] neos-3355323-arnon [MIPLIB] neos-3211096-shag [MIPLIB] neos-3603137-hoteo [MIPLIB]
Rank / ISS The image-based structural similarity (ISS) metric measures the Euclidean distance between the image-based feature vectors for the query instance and all model groups. A smaller ISS value indicates greater similarity.
1 / 0.256 2 / 0.362 357 / 1.527 505 / 1.652 802 / 2.197
Raw These images represent the CCM images in their raw forms (before any decomposition was applied) for the MIPLIB top 5.
lectsched-3 raw lectsched-2 raw neos-3355323-arnon raw neos-3211096-shag raw neos-3603137-hoteo raw

Instance Summary

The table below contains summary information for lectsched-1, the five most similar instances to lectsched-1 according to the MIC, and the five most similar instances to lectsched-1 according to MIPLIB 2017.

INSTANCE SUBMITTER DESCRIPTION ISS RANK
Parent Instance lectsched-1 [MIPLIB] Harald Schilly University lecture scheduling instance 0.000000 -
MIC Top 5 lectsched-3 [MIPLIB] Harald Schilly University lecture scheduling instance 0.256305 1
lectsched-2 [MIPLIB] Harald Schilly University lecture scheduling instance 0.362304 2
lectsched-5-obj [MIPLIB] Harald Schilly scheduling lectures at university - smaller subset of data with objective to minimize certain overlappings 0.364403 3
probportfolio [MIPLIB] Feng Qiu Sample average approximation formulation of a probabilistic portfolio optimization problem. Solved using ug[SCIP/spx], a distributed massively parallel version of SCIP run on 2,000 cores at the HLRN-II super computer facility. 0.681776 4
tanglegram6 [MIPLIB] Falk Hueffner The NP-hard Balanced Subgraph problem (variant of MaxCut) encoded as ILPs. Real-world instances from two applications from bioinformatics, finding monotone subsystems in gene regulatory networks (http://dx.doi.org/10.1007/s10878-009-9212-2) and finding optimal layouts of tanglegrams (http://dx.doi.org/10.1007/978-3-642-11269-0). 0.718930 5
MIPLIB Top 5 lectsched-3 [MIPLIB] Harald Schilly University lecture scheduling instance 0.256305 1
lectsched-2 [MIPLIB] Harald Schilly University lecture scheduling instance 0.362304 2
neos-3355323-arnon [MIPLIB] Jeff Linderoth (None provided) 1.527081 357
neos-3211096-shag [MIPLIB] Jeff Linderoth Solved with status infeasible by ParaSCIP in 42484 sec.\xa0It was confirmed by ParaXpress. 1.652191 505
neos-3603137-hoteo [MIPLIB] Jeff Linderoth (None provided) 2.196771 802


lectsched-1: Instance-to-Model Comparison Results

Model Group Assignment from MIPLIB: lectsched
Assigned Model Group Rank/ISS in the MIC: 2 / 1.288

MIC Top 5 Model Groups

These are the 5 model group composite (MGC) images that are most similar to the decomposed CCM image for the query instance, according to the ISS metric.

These are model group composite (MGC) images for the MIC top 5 model groups.
Model group: neos-pseudoapplication-109 Model group: lectsched Model group: map Model group: rmatr Model group: polygonpack
Name neos-pseudoapplication-109 lectsched map rmatr polygonpack
Rank / ISS The image-based structural similarity (ISS) metric measures the Euclidean distance between the image-based feature vectors for the query instance and all other instances. A smaller ISS value indicates greater similarity.
1 / 1.279 2 / 1.289 3 / 1.357 4 / 1.389 5 / 1.396

Model Group Summary

The table below contains summary information for the five most similar model groups to lectsched-1 according to the MIC.

MODEL GROUP SUBMITTER DESCRIPTION ISS RANK
MIC Top 5 neos-pseudoapplication-109 Jeff Linderoth (None provided) 1.279242 1
lectsched Harald Schilly University lecture scheduling model 1.288720 2
map Kiyan Ahmadizadeh Land parcel selection problems motivated by Red-Cockaded Woodpecker conservation problem 1.357426 3
rmatr Dmitry Krushinsky Model coming from a formulation of the p-Median problem using square cost matrices 1.389253 4
polygonpack Antonio Frangioni Given a set P of polygons, not necessarily convex, and a rectangle, we want to find the subset S of P with largest possible total area and a position every p in S so that there are no overlaps and they are all included in the rectangle. We allow a small set of rotations (0, 90, 180, 270 degrees) for every polygon. The problem is simplified w.r.t. the real application because the polygons do not have (fully encircled) "holes", which are supposedly filled-in separately, although they can have "bays". Models are saved as .lp. Model LpPackingModel_Dim means that we are trying to pack polygons taken from set ; there are currently 5 different sets, and is 7, 10 or 15. 1.395553 5