ICOR 2018

13th INTERNATIONAL CONFERENCE

on

OPERATIONS RESEARCH

Havana; March 6-9, 2018

 

Devoted to the 290th Onomastic of Universidad de La Habana

1728-2018

 

Plenary Themes: Optimization, Probability and Statistics, Mathematical Economic, Algorithms, Teaching on Operations Research

 

Organized by:

Universidad de La Habana

Université Paris 1, Panthéon-Sorbonne

 

Co sponsors:

Asociación Latinoamericana de Investigación Operativa. (ALIO)

American Statistical Association (ASA)

International Federation of Operations Research and Management Sciences (INFORMS)

Red Iberoamericana de Agrobigdata y Decision Support Systems" (Dss) para un Sector Agropecuario Sostenible

Red iberoamericana de estudios cuantitativos aplicados (RIDECA)

Sociedad Cubana de Matemática y Computación (SCMATCOM- Investigación Operacional).

Red Iberoamericana de Investigación en Modelos de Optimización y Decisión

y sus Aplicaciones (IMODE)


 

ONLINE REGISTRATION: http://icor2018.solwayscuba.com/


 

COMMITTEES


 

Program Committee

S. Allende [Chair, Havana], L. Álvarez [Havana], M.L. Baguer [Havana], J. M. Bardet (Paris 1), C. Bouza [Havana], G. Bouza [Havana], C. Campos [´Madrid], J. Cochran (Alabama), F. Cucker [Hong Kong], M. Cottrell [Chair, Paris 1], J. Daduna [Berlin], M. Davidian [N. Carolina], A. Ebrahimnejad [Qaemshahr], A. Fernández [Havana], C. Hardouin, [Paris Ouest Nanterre], D. Haughton [Bentley], S. Hernández [Xalapa], R. Keller[California], S. K. Roy [Vidyasagar], M. A. León [Pinar del Río], J. A. Marmolejo-Martín [Melilla], A. Marrero [Havana]], G. Molenberghs [Leuven], M. Negreiros [Fortaleza], M. Nicado [Havana], M. Olteanu (Paris 1), L. Pla [Lleida], J. Randon Furling (Paris 1), P. Ribereau [Lyon], A. Ruiz [Havana], J. Ruckmann [Bergen], J. Rynkiewicz [Paris1], L. Sandoval [Puebla], V. Sistachs [Havana], Z. Shkedy [Hasselt], R. Stolletz [Mannheim], Ch.Tammer [Halle], M. D. Ugarte [Navarra], J.L. Verdegay [Granada], B.-A. Wickström [Budapest], D. P. Wiens [Edmonton]

 

 

 

Organizing Committee

Prof. Dr. Raíl Guinovart: Dean, Facultad de Matemática y Computación [Universidad de la Habana], Danae Pérez Arjona [Universidad de la Habana], Genry Pérez Rodríguez [Universidad de la Habana].

CHRISTIANE TAMMER: On some methods to derive necessary and sufficient optimality conditions in vector optimization


 

TIMETABLE

 

Tuesday 6

9:00-14:00: Registration. Varona Building, University Campus

10:00-11:30: Opening Session, Aula Magna, University Campus.

14:00-15:00 Meeting of SCMC-ASCIO, (Cuban Association of Operations Research).

15:00-15:50: Poster Session 1.

15:50-16:10: Coffee Break

16:10-17:00: Poster Session 2.

 

Poster 1 Algorithms 1.

Chair: M. Sauto.

Blanco, Y.

A new method for decision making with the use of simulation techniques.

Campbell, J.

Evaluating an heuristic for mixed Chinese postman problems

De Armas, L.

Analysis of free Informatics tools and of open code for optimizing the problems of piling containers.

Morales, W.

Restriction-based interpolation with cubic A-splines.

Porras, C.

Multi-coverage dynamic maximal covering location problem.

Rentería, R.

Complex adaptive system in management of hospitality logistics in Colombia.

Rodríguez, N.

Alternatives of statistical analysis in parametric and non-parametric variance analysis models.

 

Poster 2: Studies on Big Data.

Chair: A. Santiago.

Castellanos, J.

Valuation Cardiovascular risk using classification models

Lao, Y.

Multi-objective model for the management of the physical resources of the logistic system in trading enterprises.

Rentería, R.

Differential morbi-mortality of victims of internal conflict and population under poverty in the Risaralda province, Colombia.

Roura, P.

Probabilistic model analysis to characterize the maximun annual wind regime

Salgado, L.

A proposal of a procedure for characterizing and developing a quantitative analysis of the labor market in Tlaxcala

Vázquez, D.E.

Data mining in the analysis of victimization and violence in Mexico.

Viada, C.

Validation of QLQ-C30 quality of life survey for patients with cancer of the head and neck, cervix, breast or prostate.

 

Wednesday 6

Morning

9:00-10:00: Tammer, Ch.: On some methods to derive necessary and sufficient optimality conditions in vector optimization Plenary Talk. Chair: G. Gfrerer.

10:00-10:30: Coffee Break

10:30-11:30: Hardouin, C.: Two-scale spatial model for binary data. Plenary Talk. Chair: J. Rynkiewicz.

11:30-13:30: Lunch

Afternoon

Room

Room 1

Room 2

Room 3

13:30.-14:50

Mathematical Economy

Optimization 1

Stochastic models 1

14:50-15:10

Coffee Break

Coffee Break

Coffee Break

15:10-16:50

Social Sciences Modeling

Algorithms 2

Stochastic models 2

RIDECA meeting

Mathematical Economy.

Chair: J. Daduna.

Time

Speaker

Title

13:30-14:10

Ahlheim, Michael

Main Talk

Economic appraisal of improved ecosystem services accruing from a reforestation project in southwest China- a multi-stakeholder approach.

14:10-14:50

Wickstroem , Bernt A

Main Talk

Informal trading networks: imitation, habits, and social evolution.

 

 

Optimization 1.

Chair C. Tammer.

Time

Speaker

Title

13:30-14:10

Gfrerer , H.

Main Talk

On a piecewise programming approach for disjunctive programming with strong convergence properties

14:10-14:30

Neruda, R.

search and optimization methods for hyper-parameter recommendation

14:30-14:50

Bouza, G.

Newton-like method for variable ordering optimization problems

 

 

Stochastic models 1.

Chair: M. Cotrell.

Time

Speaker

Title

13:30-14:10

Rynkiewicz, J.

Main Talk

On mixture of Poisson autoregressions

14:10-14:30

El Methni, J.

Kernel estimation of extreme regression risk measures

14:30-14:50

Hodgess, E.

Aggregation moving average component limits

 

 

Social Sciences Modeling.

Chair: L. Salgado.

Time

Speaker

Title

15:10-15:30

Caamal, I.

Growth rates of economic variables of the production and trade of mango in Mexico

15:30-15:50

García , J.F.

Analysis of the existent relation between innovation and labour productivity in the manufacturing industry in México. Year 2013.

15:50-16:10

Zetina, C.D.

Gender and environment of the birth place in the scientific alphabetization of students of a public Mexican university

16:10-... RIDECA Meeting.

 

 

Algorithms 2

Chair: M. Baguer.

Time

Speaker

Title

15:10-15:30

Fonseca, Y.

An improvement of reinforcement learning approach to permutational flow shop scheduling problem

15:30-15:50

Pino, J.

Automatic classification of hails in digital images

15:50-16:10

Valdés, D.

Preprocessing and segmentation of diabetic foot ulcers in Cuban patients

16:10-16:30

Pérez, C.

Solving the CVRP with infinitely many neighborhood criteria

 

Stochastic models 2.

Chair: C. Bouza.

Time

Speaker

Title

15:10-15:30

Santiago, A.

Randomized estimation a proportion using ranked set sampling and Warner`s procedure

15:30-15:50

Toledo, Á.

An application of copula theory in classification of bivariate random variables

15:50-16:10

Uranga, R.

Impact of the use of the Gibbs simulator in multiple imputation when the missingness indicator is included as a covariate

 

 

Thursday 7

Morning

9:00-10:00: Corneli, M.: Stochastic textual block modeling in dynamic networks. Plenary Talk.

Chair V. Sistachs.

10:00-10:30: Coffee Break

10:30-11:30: Cruz, C.: Definition of the soft computing concept. Description of its main components. Concept evolution since 1990 to the present day. Characterization of combinatorial problems to solve. Plenary Talk. Chair S. Allende.

11:30-13:30: Lunch.

Afternoon

Room

Room 1

Room 2

Room 3

13:30.-14:50

Spatial Statistics 1

Operations Research 1

Studies on Big Data 2

14:50-15:10

Coffee Break

Coffee Break

Coffee Break

15:10-16:50

Spatial Statistics 2

Operations Research 2

Alumni Meeting

 

Spatial Statistics 1.

Chair: P. Ribereau.

Time

Speaker

Title

13:30-13:50

Usseglio-Carleve, A.

Estimation of conditional extreme quantiles from heavy-tailed elliptical random vectors

13:50-14:10

Abu-Awwad, A.F.

A model-free selection criterion for the mixing coefficient of spatial max-mixture models.

14:30-14:50

Rentería, R.

The behavior of the homicide in a Colombian Andean city. A spatio-temporal analysis of crime through topological metrics of

 

Operations Research 1

Chair: F. García.

Time

Speaker

Title

13:30-13:50

Gaviño,G.

A system for teaching of control and monitoring of variables through virtual environments of teaching learning (EVEA) in CUUAEM Valle de México

13:50-14:10

Juárez, L. H.

Penalized and augmented Lagrangian methods for demand estimation in transit networks

14:10-14:30

González, O.

Design, construction and study of the efficiency of green panels with a monitoring system for physical variables

14:30-14:50

Pérez, V.

Goal programming synthetic index for measuring tourism destinations competitiveness

Studies on Big Data 2

Chair: C. Hardouin.

Time

Title

Talk

13:30-14:10

Lamirel, J.Ch.

Main Talk

Clustering quality based on features salience: an efficient approach

14:10-14:30

Bouza, C.N.

How big the sample data should be for accepting the normality of the sample mean?

14:30-14:50

Vázquez, Y.

Computer system for the analysis of users in Moodle

14:50-15:10

Morales, S. J.

Classification of texts based on the context by applying the model of voting algorithms

 

Spatial Statistics 2

Chair: R. Rentería.

Time

Speaker

Title

15:10-15:30

Sebrango, C.R.

Parameter estimation and real-time prediction of a dengue outbreak using model averaging: the Denguert: R package

15:30-15:50

Sepúlveda, J.J.

Analysis of adoption of ICT in Colombia through complex networks

16:50-16:10

Celleri, M.

Spatial autocorrelation of cancer deaths in Ecuador

 

Operations Research 2.

Chair: G. Gaviño.

Time

Speaker

Title

15:10-15:30

Marmolejo, J. A.

A standard linear programming model for staff planning process and its resolution using R

15:30-15:50

Sackmann, D.

Warehouse operation and big data: applications of storage location assignment based on optimization and order data analysis

15:50-16:10

Uusivuori, J.

On the extraction of a non-renewable resource with amenity value

16:10-16:30

Firoz A.

Total cost bounds in solid transportation problem with probabilistic fractional cost function under varying restrictions

19:00: Conference Dinner.

 

Friday 7

9:00-10:00: Daduna, J: Impacts of additive manufacturing on supply chain structures in industry and retail trade. Plenary Talk. Chair B.A. Wickstroem.

 

10:00-10:30: Coffee Break

 

Partial Differential equations

Chair: J.B. Baillon.

Time

 

 

10:30-11:10

Baillon, J.B.

Main Talk

About new tools to solve elliptic PDE

11:10-11:30

Estrada, J.

Stability and bifurcation of linear perturbations of the Lorenz system

 

 

11:40-12:40: Ledzewicz, U.: Optimizing combination therapies in cancer: an optimal control approach. Plenary Talk. Chair. H. Schättler.

12:30-14:00: Lunch


 

PLENARY LECTURES

 

STOCHASTIC TEXTUAL BLOCK MODELLING IN DYNAMIC NETWORKS

Marco Corneli*, Charles Bouveyron**, Pierre Latouche* and Fabrice Rossi*

*Laboratoire SAMM, Universite Paris 1 Pantheon-Sorbonne, France.

**Laboratoire J.A. Dieudonne, UMR CNRS 7351 Equipe Asclepios, INRIA

Sophia-Antipolis Universite Cote d'Azur, Nice, France.

The present paper develops a probabilistic model to cluster the nodes of a dynamic graph, accounting for the content of textual edges as well as their frequency. Vertices are clustered in groups which are homogeneous both in terms of interaction frequency and discussed topics. The dynamic graph is considered stationary on a latent time interval if the proportions of topics discussed between each pair of node groups do not change in time during that interval. A classification variational expectation-maximization (C-VEM) algorithm is adopted to perform inference. A model selection criterion is also derived to select the number of node groups, time clusters and topics. Experiments on simulated data are carried out to assess the proposed methodology. We finally illustrate an application to the Enron dataset.

 

 

DEFINITION OF THE SOFT COMPUTING CONCEPT. DESCRIPTION OF ITS MAIN COMPONENTS. CONCEPT EVOLUTION SINCE 1990 TO THE PRESENT DAY. CHARACTERIZATION OF COMBINATORIAL PROBLEMS TO SOLVE.

Carlos Cruz

Universidad de Granada, Spain.

Knowing its definition and its components, different practical applications are presented to real decision and optimization problems, ranging from tools to deal with mosquitoes that transmit dengue, to maps that adapt according to the preferences of the user.

 

 

IMPACTS OF ADDITIVE MANUFACTURING ON SUPPLY CHAIN STRUCTURES IN INDUSTRY AND RETAIL TRADE

Joachim R. Daduna

Berlin School of Economics and Law, Germany.

For several years now, the importance of Additive Manufacturing (AM) in industry and retail trade has been steadily increasing, and new fields of application are being reported almost daily. Due to technological developments, the advantages of the AM are becoming more and more apparent, which will fundamentally change the economy in the future.

Key points here are the decentralization of manufacturing structures in conjunction with a customer- and demand-oriented production. This requires structural changes in the supply chains that have not occurred for decades. Thus, for example, traditional manufacturing will decline in many industrial sectors. On the one hand, this means a drastic reduction in the required transport services due to decentralized structures, while on the other hand, the demand-oriented production leads to a substantial elimination of warehousing. This results in a significant reduction in logistics costs, whereby this positive effect will be intensified by the use of autonomous vehicles and of humanoid robots.

In industrial manufacturing, AM will in many cases fundamentally change supplier structures. Many manufacturing steps no longer must be carried-out on large-scale systems at far-away locations, but can be directly linked to final assembly. With a growing individualization of products, the application will more and more come to the fore. But even more will the AM affect the production of spare parts. This is where the AM comes in, since in this area demand can only be forecasted to a limited extent in terms of volume and its distribution over time. With a spatially dislocated on-demand manufacturing not only the logistics costs are drastically reduced, but often also the waste of materials, as they are incurred in many cases in a subtractive manufacturing.

In the sector of consumer goods manufacturing, the production of goods can be shifted to the level of retail trade to a considerable extent. This can lead to a shift to the level of regional warehouses and also to the sales facilities. In the final analysis, manufacturing can also be made on demand by the consumer himself (home fabrication). This ultimately results in many cases in a de-industrialization in connection with a return to partial self-supply which are known from pre-industrial times.

The economic benefits of AM are obvious, so the implementation is expected to be on an ever wider scale over the next few years. In comparison with the traditional manufacturing processes, there are also clear ecological advantages, for example, due to reduced material consumption based on demand-oriented manufacturing as well as lower waste volumes. However, it must also be recognized that these developments will have significant negative effects on the labor market. At least in the long term, labor demand will decline significantly as the proportion of manual added value will be significantly reduced in the changed structures.

 

 

ON SOME METHODS TO DERIVE NECESSARY AND SUFFICIENT OPTIMALITY CONDITIONS IN VECTOR OPTIMIZATION

Marius Durea*, Radu Strugariu** and Christiane Tammer***

Al. I. Cuza University, Romania.

**Gh. Asachi Technical University, Romania.

***Martin-Luther-University Halle-Wittenberg, Germany.

The aim of this talk is to address new approaches, in separate ways, to necessary and, respectively, sufficient optimality conditions in constrained vector optimization. In this respect, for the necessary optimality conditions that we derive, we use a kind of vectorial penalization technique, while for the sufficient optimality conditions we make use of an appropriate scalarization method. In both cases, the approaches couple a basic technique (of penalization or scalarization, respectively) with several results in variational analysis and optimization obtained by the authors in the last years. These combinations allow us to arrive to optimality conditions which are, in terms of assumptions made, new.

 

 

TWO-SCALE SPATIAL MODEL FOR BINARY DATA

Cécile Hardouin*, Noel Cressie**

*University Paris Nanterre, France

**University of Wollongong, Australia

A spatial lattice model for binary data is constructed from two spatial scales linked through conditional probabilities. A coarse grid of lattice locations is specified and all remaining locations (which we call the background) capture fine-scale spatial dependence. Binary data on the coarse grid are modeled with an auto-logistic distribution, conditional on the binary process on the background. The background behavior is captured through a hidden Gaussian process after a logit transformation on its Bernoulli success probabilities. The likelihood is then the product of the (conditional) autologistic probability distribution and the hidden Gaussian—Bernoulli process. The parameters of the new model come from both spatial scales. A series of simulations illustrates the spatial-dependence properties of the model and likelihood-based methods are used to estimate its parameters.

 

 

OPTIMIZING COMBINATION THERAPIES IN CANCER: AN OPTIMAL CONTROL APPROACH

Urszula Ledzewicz* and Heinz Schättler**

*Southern Illinois University Edwardsville, USA and Lodz University of Technology, Poland.

**Washington University, USA.

Because of the complexities of cancer growth and its interactions with the tumor microenvironment, modern cancer treatment protocols are multi-targeted and take into account highly diverse subpopulations of cancerous cells with widely varying therapeutic sensitivities all embedded into the tumor microenvironment. This includes the vasculature as well as the elements of the immune system. Given this complex scenario, dosage, frequency and sequencing of therapeutic agents may have a major effect on the outcome of treatment. Thus, the questions “how much, how often, in what sequence” the anticancer drugs should be given to secure an optimal outcome are essential and far from being answered. Simple answers are not always the best ones. Actually, there is mounting medical evidence that "more is not necessarily better" and a properly calibrated dose which takes into account this complexity can lead to a better outcome.

Formulating mathematical models with an objective that reflects the overall goal of the therapy, like minimizing the tumor size and side effects, maximizing the actions of the immune system etc., leads to optimal control problems where mathematical analysis can answer some of these questions in a theoretical framework. Using the Pontryagin maximum principle and tools of geometric optimal control theory insights can be gained into the optimality of bang-bang controls (representing medically maximum tolerated doses, MTD) and singular controls (corresponding to biologically optimal doses, BOD). We present results about models for combinations of anti-angiogenic inhibitors with chemo- and radiotherapy as well as models including tumor-immune system interactions.