Djamila Hamdadou

Towards a New Multicriteria Decision Support Method Using Fuzzy Measures and the Choquet Integral

In literature, there is a large panoply of multicriteria analysis methods MCAM, each one is characterized by the nature of its input data, the way to edit its outputs and the operations used to perform calculations especially the performances aggregation. Aggregation is the operation consisting in grouping several quantities in a unique value in order to facilitate the manipulation and the interpretation of the original values. MCAM are classified according to the type of aggregation that they perform, so we can distinguish total, partial and local aggregation. Each MCAM has advantages and suffers from some limits. In this paper, the authors proposed a new multicriteria analysis method AMFI dedicated to solve ranking decision support problems. AMFI is based on the use of fuzzy measures and Choquet integral to represent interactions between criteria and improve the coherence of the results. The authors proceeded to a series of experimentations allowing highlighting theoretical elements of the proposed method and they performed sensitivity analysis to test its robustness.

Une nouvelle approche pour la régulation et la reconfiguration spatiale d’un réseau de transport urbain collectif

This article proposes the integration of an approach based on cellular automata for the regulation and the reconfiguration of urban transportation systems. The complexity of an UTS has a dramatic effect on disturbances. Incidents are very complicated when there are several areas particularly concerned or blocking a route for a longer or shorter. In this case, appropriate regulation action or any particular route changes are necessary. We focus in this paper to integrate this new approach in an UTS regulating platform SARRT to improve collective performance point of view response time and data storage. To implement this approach we use a Boolean Modeling Language (BML) adopted by the cellular inference engine (CIE).

SEIR-SW, Simulation Model of Influenza Spread Based on the Small World Network

This study modeled the spread of an influenza epidemic in the population of Oran, Algeria. We investigated the mathematical epidemic model, SEIR (Susceptible-Exposed-Infected-Removed), through extensive simulations of the effects of social network on epidemic spread in a Small World (SW) network, to understand how an influenza epidemic spreads through a human population. A combined SEIR-SW model was built, to help understand the dynamics of infectious disease in a community, and to identify the main characteristics of epidemic transmission and its evolution over time. The model was also used to examine social network effects to better understand the topological structure of social contact and the impact of its properties. Experiments were conducted to evaluate the combined SEIR-SW model. Simulation results were analyzed to explore how network evolution influences the spread of desease, and statistical tests were applied to validate the model. The model accurately replicated the dynamic behavior of the real influenza epidemic data, confirming that the susceptible size and topological structure of social networks in a human population significantly influence the spread of infectious diseases. Our model can provide health policy decision makers with a better understanding of epidemic spread, allowing them to implement control measures. It also provides an early warning of the emergence of influenza epidemics.

The MAV-ES Data Integration Approach for Decisional Information Systems DIS: A Case on Epidemiologic Monitoring

Massive data to facilitate decision making for organizations and their corporate users exist in many forms, types and formats. Importantly, the acquisition and retrieval of relevant supporting information should be timely, precise and complete. Unfortunately, due to differences in syntax and semantics, the extraction and integration of available semi-structured data from different sources often fail. Needs for seamless and effective data integration so as to access, retrieve and use information from diverse data sources cannot be overly emphasized. Moreover, information external to organizations may also often have to be sourced for the intended users through a smart data integration system. Owing to the open, dynamic and heterogeneity nature of data, data integration is becoming an increasingly complex process. A new data integration approach encapsulating mediator systems and data warehouse is proposed here. Aside from the heterogeneity of data sources, other data integration design problems include distinguishing the definition of the global schema, the mappings and query processing. In order to meet all of these challenges, the authors of this paper advocate an approach named MAV-ES, which is characterized by an architecture based on a global schema, partial schemas and a set of sources. The primary benefit of this architecture is that it combines the two basic GAV and LAV approaches so as to realize added-value benefits of the mixed approach.