Elisa Schaeffer | McGill University (original) (raw)

Papers by Elisa Schaeffer

Proceedings of MOL2NET'22, Conference on Molecular, Biomedical & Computational Sciences and Engineering, 8th ed. - MOL2NET: FROM MOLECULES TO NETWORKS

The application of multicriteria analysis methods to the creation of models that allow evaluating... more The application of multicriteria analysis methods to the creation of models that allow evaluating the pleasure of the smoker, and optimizing the integral quality of the tobacco leaf, ensuring the degree of sensory satisfaction and the harmfulness of the habit, is presented. The quality model was verified in practice, comparing it with

Physica A: Statistical Mechanics and its Applications

In this paper, we present a new algorithm to route text queries within a P2P network, called Neig... more In this paper, we present a new algorithm to route text queries within a P2P network, called Neighboring-Ant Search (NAS) algorithm. The algorithm is based on the Ant Colony System metaheuristic and the SemAnt algorithm. More so, NAS is hybridized with local environment strategies of learning, characterization, and exploration. Two Learning Rules (LR) are used to learn from past performance, these rules are modified by three new Learning Functions (LF). A Degree-Dispersion-Coefficient (DDC) as a local topological metric is used for the structural characterization. A variant of the well-known one-step Lookahead exploration is used to search the nearby environment. These local strategies make NAS self-adaptive and improve the performance of the distributed search. Our results show the contribution of each proposed strategy to the performance of the NAS algorithm. The results reveal that NAS algorithm outperforms methods proposed in the literature, such as Random-Walk and SemAnt.

All in-text references underlined in blue are linked to publications on ResearchGate, letting you... more All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.

In this work, we address the problem of locally estimating the size of a Peerto-Peer (P2P) networ... more In this work, we address the problem of locally estimating the size of a Peerto-Peer (P2P) network using local information. We present a novel approach for estimating the size of a peer-to-peer (P2P) network, fitting the sum of new neighbors discovered at each iteration of a breadth-first search (BFS) with a logarithmic function, and then using Lambert’s W function to solve a root of a ln(n)+b−n = 0, where n is the network size. With rather little computation, we reach an estimation error of at most 10 percent, only allowing the BFS to iterate to the third level.

Bulletin of the American Physical Society, 2015

The Prize-collecting Rural Postman Problem, also known as the Privatized Rural Postman Problem, i... more The Prize-collecting Rural Postman Problem, also known as the Privatized Rural Postman Problem, is an arc-routing problem where each demand edge is associated with a profit, which is collected once if the edge is served, independently of the number of traversals. Included edges incur in routing costs proportionally to the number of traversals. In this paper, we introduce the Windy Privatized Rural Postman Problem in which routing costs also depend on the direction of the traversals. For this problem, we propose a solution heuristic based on ant-colony optimization. The proposed method is capable of constructing profitable closed walks with low computational load. The quality of the obtained solutions can be assessed comparing their values to lower bounds.

EN ESTE TRABAJO SE PRESENTA UN MODELO DE PROGRAMACION LINEAL ENTERA MIXTA (MILP) PARA EL PROBLEMA... more EN ESTE TRABAJO SE PRESENTA UN MODELO DE PROGRAMACION LINEAL ENTERA MIXTA (MILP) PARA EL PROBLEMA DEL PORTAFOLIO DE PROYECTOS PUBLICOS R&D BI-OBJETIVO. EL ENFOQUE PROPUESTO PROVEE UN PUNTO MEDIO ENTRE EL IMPACTO Y EL NUMERO DE LOS PROYECTOS. SE CONSIDERAN TECNICAS DE RELAJACION LAGRANGIANA PARA OBTENER COTAS FACILMENTE CALCULABLES PARA LOS VALORES OBJETIVOS. LA EXPERIMENTACION MUESTRA QUE PUEDE OBTENERSE UNA SOLUCION EN MENOS DE UN MINUTO INCLUSO PARA CASOS DE CARTERAS DE MAS DE 25,000 PROYECTOS PROPUESTOS. ESTO IMPLICA UNA MEJORA SIGNIFICATIVA A LOS ENFOQUES PREVIOS QUE RESUELVEN EFICIENTEMENTE CASOS CON SOLO ALGUNOS CIENTOS DE PROYECTOS

In this paper, we present a new algorithm to route text queries within a P2P network, called Neig... more In this paper, we present a new algorithm to route text queries within a P2P network, called Neighboring-Ant Search (NAS) algorithm. The algorithm is based on the Ant Colony System metaheuristic and the SemAnt algorithm. More so, NAS is hybridized with local environment strategies of learning, characterization, and exploration. Two Learning Rules (LR) are used to learn from past performance, these rules are modified by three new Learning Functions (LF). A Degree-Dispersion-Coefficient (DDC) as a local topological metric is used for the structural characterization. A variant of the well-known one-step Lookahead exploration is used to search the nearby environment. These local strategies make NAS self-adaptive and improve the performance of the distributed search. Our results show the contribution of each proposed strategy to the performance of the NAS algorithm. The results reveal that NAS algorithm outperforms methods proposed in the literature, such as Random-Walk and SemAnt.

En este trabajo se presenta un metodo de agrupamiento local en grafos dirigidos: dado un vertice ... more En este trabajo se presenta un metodo de agrupamiento local en grafos dirigidos: dado un vertice semilla, se determino el grupo de vertices al que pertenece de tal forma que los vertices seleccionados sean estructuralmente cercanos de la semilla. A un grafo dirigido se le puede asociar una cadena de Markov que corresponde a una caminata aleatoria ciega en el grafo. Se aprovecho esta conexion para expresar cercania estructural en terminos de los tiempos de absorcion para detectar vertices que son “cercanos” al vertice semilla. Se detecto el grupo de un vertice a traves de caminatas aleatorias cortas repetidas desde el vertice semilla, analizando la frecuencia de visitas a los otros vertices. Se experimento con grafos pequenos para comparar el resultado los tiempos exactos de absorcion. El agrupamiento local puede ser aplicado a diferentes fenomenos reales, por ejemplo, en propagacion de epidemias, balanceo de carga, etc

2017 Computing Conference

Classical computational complexity studies the asymptotic relationship between instance size and ... more Classical computational complexity studies the asymptotic relationship between instance size and the amount of resources consumed in the worst case. However, it has become evident that the instance size by itself is an insufficient measure and that the worst-case scenario is often uninformative in practice. As a complementary analysis, the examination of structural properties present in the instances and the effects they have on algorithm performance are proposed; the goal is to characterize complexity in terms of instance structure. A framework for identifying and characterizing hard instances based on algorithm behaviour is proposed as well as a case study applying the framework on the graph coloring problem is presented.

The Minerals, Metals & Materials Series, 2017

We seek to explain the weld toe-crack failure in terms of crack nucleation and propagation based ... more We seek to explain the weld toe-crack failure in terms of crack nucleation and propagation based on the observed phases and their size. Three welded joins were studied, varying the heat inputs in order to modify the weld microstructure. We work with an API X70 PSL2 pipeline steel, varying the ferritic size phases. We performed Charpy impact and Vickers hardness tests; we also determined the chemical composition (especially the oxygen percentage). Based on the relationship between the test results and the microstructure, we achieve a modification of the welded join microstructure and toughness, observing the heat-input effect in the weld oxygen diffusion as well as its effect in the type and size of the phase observed.

Journal of Automation, Mobile Robotics and Intelligent Systems, 2011

Bulletin of the American Physical Society, Mar 3, 2015

It has become common for internet users to form online communities to share resources and informa... more It has become common for internet users to form online communities to share resources and information. However, the development of such virtual communities suffers from the lack of efficient search methods. Among the many issues that make search of distributed resources a complex task, is dynamic nature of both the connection topology and the resources contained in the network of community members. The Internet in general is considered a complex network for its size and interconnectivity that both constantly evolve. In complex dynamic distributed systems, global knowledge recollection is not a feasible approach. In this work we present a study of two algorithms for distributed search of text information (Semantic Query Routing), enriched with a local topology-characterization metric, seeking to improve the performance of distributed search. Our results show that including such a metric in the Neighboring-Ant-Search algorithm improves its performance 50% in terms of the number of hops needed to locate a set of resources in scale-free and small-world networks.

Computacion Y Sistemas Num 4 Vol 13, 2014

Siam Journal on Scientific Computing, 2010

We address the problem of obtaining by local computation a sample of a graph such that the vertic... more We address the problem of obtaining by local computation a sample of a graph such that the vertices are sampled uniformly. The presented solution uses a Markov chain to combine a rapidly mixing random walk that does not reach a uniform distribution with a slow-mixing walk that does. The resulting chain mixes at a notably faster rate than the standard uniform random walk and can be simulated exactly using only local information.

We propose efficient algorithms for two key tasks in the analysis of large nonuniform networks: u... more We propose efficient algorithms for two key tasks in the analysis of large nonuniform networks: uniform node sampling and cluster detection. Our sampling technique is based on augmenting a simple, but slowly mixing uniform MCMC sampler with a regular random walk in order to speed up its convergence; however the combined MCMC chain is then only sampled when it is

Computacion Y Sistemas, Jun 1, 2010

Proceedings of MOL2NET'22, Conference on Molecular, Biomedical & Computational Sciences and Engineering, 8th ed. - MOL2NET: FROM MOLECULES TO NETWORKS

The application of multicriteria analysis methods to the creation of models that allow evaluating... more The application of multicriteria analysis methods to the creation of models that allow evaluating the pleasure of the smoker, and optimizing the integral quality of the tobacco leaf, ensuring the degree of sensory satisfaction and the harmfulness of the habit, is presented. The quality model was verified in practice, comparing it with

Physica A: Statistical Mechanics and its Applications

In this paper, we present a new algorithm to route text queries within a P2P network, called Neig... more In this paper, we present a new algorithm to route text queries within a P2P network, called Neighboring-Ant Search (NAS) algorithm. The algorithm is based on the Ant Colony System metaheuristic and the SemAnt algorithm. More so, NAS is hybridized with local environment strategies of learning, characterization, and exploration. Two Learning Rules (LR) are used to learn from past performance, these rules are modified by three new Learning Functions (LF). A Degree-Dispersion-Coefficient (DDC) as a local topological metric is used for the structural characterization. A variant of the well-known one-step Lookahead exploration is used to search the nearby environment. These local strategies make NAS self-adaptive and improve the performance of the distributed search. Our results show the contribution of each proposed strategy to the performance of the NAS algorithm. The results reveal that NAS algorithm outperforms methods proposed in the literature, such as Random-Walk and SemAnt.

All in-text references underlined in blue are linked to publications on ResearchGate, letting you... more All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.

In this work, we address the problem of locally estimating the size of a Peerto-Peer (P2P) networ... more In this work, we address the problem of locally estimating the size of a Peerto-Peer (P2P) network using local information. We present a novel approach for estimating the size of a peer-to-peer (P2P) network, fitting the sum of new neighbors discovered at each iteration of a breadth-first search (BFS) with a logarithmic function, and then using Lambert’s W function to solve a root of a ln(n)+b−n = 0, where n is the network size. With rather little computation, we reach an estimation error of at most 10 percent, only allowing the BFS to iterate to the third level.

Bulletin of the American Physical Society, 2015

The Prize-collecting Rural Postman Problem, also known as the Privatized Rural Postman Problem, i... more The Prize-collecting Rural Postman Problem, also known as the Privatized Rural Postman Problem, is an arc-routing problem where each demand edge is associated with a profit, which is collected once if the edge is served, independently of the number of traversals. Included edges incur in routing costs proportionally to the number of traversals. In this paper, we introduce the Windy Privatized Rural Postman Problem in which routing costs also depend on the direction of the traversals. For this problem, we propose a solution heuristic based on ant-colony optimization. The proposed method is capable of constructing profitable closed walks with low computational load. The quality of the obtained solutions can be assessed comparing their values to lower bounds.

EN ESTE TRABAJO SE PRESENTA UN MODELO DE PROGRAMACION LINEAL ENTERA MIXTA (MILP) PARA EL PROBLEMA... more EN ESTE TRABAJO SE PRESENTA UN MODELO DE PROGRAMACION LINEAL ENTERA MIXTA (MILP) PARA EL PROBLEMA DEL PORTAFOLIO DE PROYECTOS PUBLICOS R&D BI-OBJETIVO. EL ENFOQUE PROPUESTO PROVEE UN PUNTO MEDIO ENTRE EL IMPACTO Y EL NUMERO DE LOS PROYECTOS. SE CONSIDERAN TECNICAS DE RELAJACION LAGRANGIANA PARA OBTENER COTAS FACILMENTE CALCULABLES PARA LOS VALORES OBJETIVOS. LA EXPERIMENTACION MUESTRA QUE PUEDE OBTENERSE UNA SOLUCION EN MENOS DE UN MINUTO INCLUSO PARA CASOS DE CARTERAS DE MAS DE 25,000 PROYECTOS PROPUESTOS. ESTO IMPLICA UNA MEJORA SIGNIFICATIVA A LOS ENFOQUES PREVIOS QUE RESUELVEN EFICIENTEMENTE CASOS CON SOLO ALGUNOS CIENTOS DE PROYECTOS

In this paper, we present a new algorithm to route text queries within a P2P network, called Neig... more In this paper, we present a new algorithm to route text queries within a P2P network, called Neighboring-Ant Search (NAS) algorithm. The algorithm is based on the Ant Colony System metaheuristic and the SemAnt algorithm. More so, NAS is hybridized with local environment strategies of learning, characterization, and exploration. Two Learning Rules (LR) are used to learn from past performance, these rules are modified by three new Learning Functions (LF). A Degree-Dispersion-Coefficient (DDC) as a local topological metric is used for the structural characterization. A variant of the well-known one-step Lookahead exploration is used to search the nearby environment. These local strategies make NAS self-adaptive and improve the performance of the distributed search. Our results show the contribution of each proposed strategy to the performance of the NAS algorithm. The results reveal that NAS algorithm outperforms methods proposed in the literature, such as Random-Walk and SemAnt.

En este trabajo se presenta un metodo de agrupamiento local en grafos dirigidos: dado un vertice ... more En este trabajo se presenta un metodo de agrupamiento local en grafos dirigidos: dado un vertice semilla, se determino el grupo de vertices al que pertenece de tal forma que los vertices seleccionados sean estructuralmente cercanos de la semilla. A un grafo dirigido se le puede asociar una cadena de Markov que corresponde a una caminata aleatoria ciega en el grafo. Se aprovecho esta conexion para expresar cercania estructural en terminos de los tiempos de absorcion para detectar vertices que son “cercanos” al vertice semilla. Se detecto el grupo de un vertice a traves de caminatas aleatorias cortas repetidas desde el vertice semilla, analizando la frecuencia de visitas a los otros vertices. Se experimento con grafos pequenos para comparar el resultado los tiempos exactos de absorcion. El agrupamiento local puede ser aplicado a diferentes fenomenos reales, por ejemplo, en propagacion de epidemias, balanceo de carga, etc

2017 Computing Conference

Classical computational complexity studies the asymptotic relationship between instance size and ... more Classical computational complexity studies the asymptotic relationship between instance size and the amount of resources consumed in the worst case. However, it has become evident that the instance size by itself is an insufficient measure and that the worst-case scenario is often uninformative in practice. As a complementary analysis, the examination of structural properties present in the instances and the effects they have on algorithm performance are proposed; the goal is to characterize complexity in terms of instance structure. A framework for identifying and characterizing hard instances based on algorithm behaviour is proposed as well as a case study applying the framework on the graph coloring problem is presented.

The Minerals, Metals & Materials Series, 2017

We seek to explain the weld toe-crack failure in terms of crack nucleation and propagation based ... more We seek to explain the weld toe-crack failure in terms of crack nucleation and propagation based on the observed phases and their size. Three welded joins were studied, varying the heat inputs in order to modify the weld microstructure. We work with an API X70 PSL2 pipeline steel, varying the ferritic size phases. We performed Charpy impact and Vickers hardness tests; we also determined the chemical composition (especially the oxygen percentage). Based on the relationship between the test results and the microstructure, we achieve a modification of the welded join microstructure and toughness, observing the heat-input effect in the weld oxygen diffusion as well as its effect in the type and size of the phase observed.

Journal of Automation, Mobile Robotics and Intelligent Systems, 2011

Bulletin of the American Physical Society, Mar 3, 2015

It has become common for internet users to form online communities to share resources and informa... more It has become common for internet users to form online communities to share resources and information. However, the development of such virtual communities suffers from the lack of efficient search methods. Among the many issues that make search of distributed resources a complex task, is dynamic nature of both the connection topology and the resources contained in the network of community members. The Internet in general is considered a complex network for its size and interconnectivity that both constantly evolve. In complex dynamic distributed systems, global knowledge recollection is not a feasible approach. In this work we present a study of two algorithms for distributed search of text information (Semantic Query Routing), enriched with a local topology-characterization metric, seeking to improve the performance of distributed search. Our results show that including such a metric in the Neighboring-Ant-Search algorithm improves its performance 50% in terms of the number of hops needed to locate a set of resources in scale-free and small-world networks.

Computacion Y Sistemas Num 4 Vol 13, 2014

Siam Journal on Scientific Computing, 2010

We address the problem of obtaining by local computation a sample of a graph such that the vertic... more We address the problem of obtaining by local computation a sample of a graph such that the vertices are sampled uniformly. The presented solution uses a Markov chain to combine a rapidly mixing random walk that does not reach a uniform distribution with a slow-mixing walk that does. The resulting chain mixes at a notably faster rate than the standard uniform random walk and can be simulated exactly using only local information.

We propose efficient algorithms for two key tasks in the analysis of large nonuniform networks: u... more We propose efficient algorithms for two key tasks in the analysis of large nonuniform networks: uniform node sampling and cluster detection. Our sampling technique is based on augmenting a simple, but slowly mixing uniform MCMC sampler with a regular random walk in order to speed up its convergence; however the combined MCMC chain is then only sampled when it is

Computacion Y Sistemas, Jun 1, 2010