Offre de th
Forum 'Emplois' - Sujet créé le 2015-03-09 par Vincent Tkindt
Title: Contributions of operations research to the inexact matching problem
Key points :
* Topic: This Ph.D. thesis aims at elaborating models and methods to solve the error-tolerant matching problem.
* Key-words: Operational research, continuous / discrete optimization and graph matching.
* Proposal: Ph.D. thesis proposal in Computer Science. Field of science: Operational Research
* Skills: Mathematical programming, combinatorial optimization, exact
methods, heuristics and C++/Java programming.
* Funding: Doctoral contract funding by Region Centre. Gross salary:
1600 euros / month.
* Advisors: Romain Raveaux, Vincent T'Kindt and Donatello Conte.
* Organization: LI laboratory, Université de Tours.
* Where: Tours, France.
* Deadline to apply: Until the 7th of May 2015.
* Start date: preferably in September / October 2015.
* Duration: 3 years
* Contact: romain.raveaux@univ-tours.fr
Short summary:
Attributed graphs provide convenient structures for representing objects when relational properties are of interest. Such representations are frequently useful in applications ranging from computer-aided drug design to machine vision. A familiar machine vision problem is to recognize specific objects within an image.
In this case, the image is processed to generate a representative graph based on structural characteristics, such a region adjacency graph or a line adjacency graph, and vertex attributes may be assigned according to characteristics of the region to which each vertex corresponds. For attributed graphs, both vertices and edges can be characterized by attributes that can vary from nominal labels to more complex descriptions such as strings or features vectors, leading to very powerful representations. This representative graph is then compared to a database of prototype or model graphs in order to identify and classify the object of interest. Graph matching (GM) plays a central role in many computer science problems, ranging from shape matching in 2-D and 3-D, object categorization, feature tracking, symmetry analysis, action recognition to protein alignment. In this context, a reliable and speedy method for comparing graphs is important. To this aim, combinatorial optimization and mathematical programming are key tools to propose exact methods. In addition, heuristics will be investigated based on discrete and continuous optimization.
How to apply ?
Send a CV + motivation letter + your master transcripts to Romain Raveaux (romain.raveaux@univ-tours.fr).
A more detail description of the subject can be also provided.
Key points :
* Topic: This Ph.D. thesis aims at elaborating models and methods to solve the error-tolerant matching problem.
* Key-words: Operational research, continuous / discrete optimization and graph matching.
* Proposal: Ph.D. thesis proposal in Computer Science. Field of science: Operational Research
* Skills: Mathematical programming, combinatorial optimization, exact
methods, heuristics and C++/Java programming.
* Funding: Doctoral contract funding by Region Centre. Gross salary:
1600 euros / month.
* Advisors: Romain Raveaux, Vincent T'Kindt and Donatello Conte.
* Organization: LI laboratory, Université de Tours.
* Where: Tours, France.
* Deadline to apply: Until the 7th of May 2015.
* Start date: preferably in September / October 2015.
* Duration: 3 years
* Contact: romain.raveaux@univ-tours.fr
Short summary:
Attributed graphs provide convenient structures for representing objects when relational properties are of interest. Such representations are frequently useful in applications ranging from computer-aided drug design to machine vision. A familiar machine vision problem is to recognize specific objects within an image.
In this case, the image is processed to generate a representative graph based on structural characteristics, such a region adjacency graph or a line adjacency graph, and vertex attributes may be assigned according to characteristics of the region to which each vertex corresponds. For attributed graphs, both vertices and edges can be characterized by attributes that can vary from nominal labels to more complex descriptions such as strings or features vectors, leading to very powerful representations. This representative graph is then compared to a database of prototype or model graphs in order to identify and classify the object of interest. Graph matching (GM) plays a central role in many computer science problems, ranging from shape matching in 2-D and 3-D, object categorization, feature tracking, symmetry analysis, action recognition to protein alignment. In this context, a reliable and speedy method for comparing graphs is important. To this aim, combinatorial optimization and mathematical programming are key tools to propose exact methods. In addition, heuristics will be investigated based on discrete and continuous optimization.
How to apply ?
Send a CV + motivation letter + your master transcripts to Romain Raveaux (romain.raveaux@univ-tours.fr).
A more detail description of the subject can be also provided.