Marcos L. P. Bueno

Multiobjective Evolutionary Algorithms and a Combined Heuristic for Route Reconnection Applied to Multicast Flow Routing

Multicast transmission corresponds to send data to several destinations often involving requirements of Quality of Service (QoS) and Traffic Engineering (TE). This work investigates new evolutionary models to tackle Multicast Flow Routing in a Pareto multiobjective perspective. Two multiobjective evolutionary algorithms (SPEA and SPEA2) were applied as the underlying search of such models. QoS and TE requirements are also considered in the route calculus by optimizing four objectives - maximum link utilization, total cost, maximum end-to-end delay and hops count - attending a link capacity constraint. Besides, three heuristics for subtrees reconnection to be used on crossover and mutation operators are investigated. The first heuristic uses a shortest path algorithm to improve the convergence; the second one employs a random search to reconnect subtrees into a valid route and the third one mixes the other two combining the good skills of each one. The resultant evolutionary environments were evaluated using four multiobjective metrics: two for convergence and two for diversity. SPEA2 reached better results than SPEA on the vast majority cases. The design of crossover and mutation operators that provide more diversity lead to very good improvements on both multiobjective goals - convergence and diversity - being that the heuristic which combines random and shortest path reconnections and returned the best results.

Algorithms to Augment Diversity and Convergence in Multiobjective Multicast Flow Routing

Multicast transmission corresponds to send data to several destinations, often involving requirements of Quality of Service (QoS) and Traffic Engineering (TE). These multiple requirements lead to the need of optimizing a set of conflicting objectives subject to constraints. Starting from the well-known evolutionary algorithm SPEA2, two formulations for the Routing problem were considered, minimizing four objectives - maximum link utilization, total cost, maximum end-to-end delay and hops count - subject to a link capacity constraint. The key investigation performed here is about the incorporation of a mechanism to reduced repeated individuals along the population, evaluating three algorithms for such task. One of them (\ftm) is proposed on this study, whose results in six instances of the problem with five metrics to evaluate convergence and diversity goals, indicates that the evolutionary model based on SPEA2 using our algorithm had returned the better results.

Asymmetric hidden Markov models

Abstract In many problems involving multivariate time series, hidden Markov models (HMMs) are often employed for modeling complex behavior over time. HMMs can, however, require large number of states, what can lead to poor problem insight and model overfitting, especially when limited data is available. In this paper, we further investigate the family of asymmetric hidden Markov models (HMM-As), which generalize the emission distributions to arbitrary Bayesian-network distributions, allowing for state-specific graphical structures in the feature space. As a consequence, HMM-As are able to render more compact state spaces, thus from a learning perspective HMM-As can better handle the complexity-overfitting trade-off. In this paper, we study representation properties of asymmetric and symmetric HMMs, as well as provide a learning algorithm for HMM-As. We provide empirical results based on simulations for comparing HMM-As with symmetric and other asymmetry-aware models, showing that modeling more general asymmetries can be very effective. We also consider real-world datasets from several domains, aiming to show that multiple graphical structures underlying data can be identified and are able to provide additional problem insight. Although learning HMM-As can be more complex, it is shown that it is feasible in practice due to their ability to maintain compact state spaces, yet more expressive ones.

Pareto-Based Optimization of Multicast Flows with QoS and Traffic Engineering Requirements

Multicast routing consists in sending information in computer networks to a selective number of destinations. QoS and Traffic Engineering requirements can also be considered in such kind of routing, leading to the need of optimizing a set of objectives subject to constraints. We investigated algorithms to perform the calculus of multicast routes while minimizing four objectives - maximum link utilization, total cost, maximum end-to-end delay and mean end-to-end delay - attending a link capacity constraint. New multiobjective evolutionary models to tackle multicast routing are discussed here based on SPEA2. Besides, two heuristics for subtrees reconnection to be used on crossover and mutation operators are investigated. The first heuristic uses a shortest path algorithm; the second one employs a random search. Our results indicate that the evolutionary model based on SPEA2 using the random search heuristic returned the best performance. The advantage of such approach is observed by comparing the routes obtained using our multiobjective environment with those returned by SPT.

Understanding disease processes by partitioned dynamic Bayesian networks

For many clinical problems in patients the underlying pathophysiological process changes in the course of time as a result of medical interventions. In model building for such problems, the typical scarcity of data in a clinical setting has been often compensated by utilizing time homogeneous models, such as dynamic Bayesian networks. As a consequence, the specificities of the underlying process are lost in the obtained models. In the current work, we propose the new concept of partitioned dynamic Bayesian networks to capture distribution regime changes, i.e. time non-homogeneity, benefiting from an intuitive and compact representation with the solid theoretical foundation of Bayesian network models. In order to balance specificity and simplicity in real-world scenarios, we propose a heuristic algorithm to search and learn these non-homogeneous models taking into account a preference for less complex models. An extensive set of experiments were ran, in which simulating experiments show that the heuristic algorithm was capable of constructing well-suited solutions, in terms of goodness of fit and statistical distance to the original distributions, in consonance with the underlying processes that generated data, whether it was homogeneous or non-homogeneous. Finally, a study case on psychotic depression was conducted using non-homogeneous models learned by the heuristic, leading to insightful answers for clinically relevant questions concerning the dynamics of this mental disorder.

MEANDS: A Many-objective Evolutionary Algorithm based on Non-dominated Decomposed Sets applied to multicast routing

Abstract Evolutionary algorithms have emerged in the last twenty years as a powerful approach for dealing with multi-objective optimization problems (MOPs). Although classical multi-objective evolutionary algorithms (MOEAs), such as SPEA2 and NSGA-II, have been designed to manipulate any number of objectives, the results of their practical application to MOPs with more than three objectives revealed that they have limitations. Many-objective algorithms represent the novelty in MOEA research because they are specially designed to handle search spaces of high dimension. In this paper, a new evolutionary algorithm able to handle discrete optimization problems with many objectives is proposed, called Many-objective Evolutionary Algorithm based on Non-dominated Decomposed Sets (MEANDS). MEANDS decomposes the original MOP into several, simpler MOPs, for which sub-populations of non-dominated solutions are maintained and evolved together. MEANDS relaxes several restrictions of predecessor algorithms, such as the size of sub-populations and the need for weights in the lower dimension MOPs. Empirical results show that MEANDS was able to find better solutions than those from well-known MOEAs (NSGA-III, SPEA2, SPEA2+SDE, MOEA/D, MOEA/DD, and its predecessor MEAMT) in the multicast routing problem involving 4, 5, and 6 QoS-based objectives.

Towards Adaptive Scheduling of Maintenance for Cyber-Physical Systems

Scheduling and control of Cyber-Physical Systems (CPS) are becoming increasingly complex, requiring the development of new techniques that can effectively lead to their advancement. This is also the case for failure detection and scheduling component replacements. The large number of factors that influence how failures occur during operation of a CPS may result in maintenance policies that are time-monitoring based, which can lead to suboptimal scheduling of maintenance. This paper investigates how to improve maintenance scheduling of such complex embedded systems, by means of monitoring in real-time the critical components and dynamically adjusting the optimal time between maintenance actions. The proposed technique relies on machine learning classification models in order to classify component failure cases vs. non-failure cases, and on real-time updating of the maintenance policy of the sub-system in question. The results obtained from the domain of printers show that a model that is responsive to the environmental changes can enable consumable savings, while keeping the same product quality, and thus be relevant for industrial purposes.

Many-Objective Evolutionary Algorithms for Multicast Routing with Quality of Service Problem

In order to guarantee a proper end-to-end communication in computer networks, many applications impose Quality of Service (QoS) requirements. In QoS routing, multiple conflicting objectives are optimized simultaneously. On the other hand, multicasting (i.e. sending data to multiple destinations) in its simplest form is a computationally difficult problem, which becomes more challenging when QoS requirements are considered. Previous works have investigated the use of multi-objective evolutionary algorithms (MOEAs) in the multicast routing with QoS problem under a Pareto-dominance approach. In particular, the so-called SPEA2 algorithm has provided good results in problem formulations involving up to four objectives. However, it has been recognized that for higher-dimension problems the convergence of MOEAs can decline substantially. In this work, we propose a model for routing based on the many-objectives algorithm MEAMT (multi-objective evolutionary algorithms with many tables), aiming at better dealing with the difficulty faced by state-of-the-art routing methods when solving for higherdimensional formulations. We present the new model for multicast routing based on MEAMT, as well as report experiments on a comparison with SPEA2-based routing in formulations involving 4, 5 and 6 objectives over several problem instances with different topologies and sizes. The results indicate that the new routing model overcomes SPEA2 in most cases, specially when solving for higher dimensional problems in routing.

A Dynamic Multiobjective Evolutionary Algorithm for Multicast Routing Problem

In this work, we propose an evolutionary algorithm to tackle a multiobjective optimization problem, namely a constrained multicast routing with quality demands. The proposed algorithm embeds two different strategies along with SPEA2 (Strength Pareto Evolutionary Algorithm 2) method attempting to improve convergence to Pareto front. These strategies are a heuristic for population diversity augmentation and a neighborhood mating selection scheme. Experimental results showed that selecting which strategy to use depends on population dynamics aspects described by non dominated solutions over evolutionary iterations. It was possible to observe that the proposed mechanism can help the algorithm to achieve better solutions over convergence and diversity goals in most cases.

Modeling the Dynamics of Multiple Disease Occurrence by Latent States.

The current availability of large volumes of health care data makes it a promising data source to new views on disease interaction. Most of the times, patients have multiple diseases instead of a single one (also known as multimorbidity), but the small size of most clinical research data makes it hard to impossible to investigate this issue. In this paper, we propose a latent-based approach to expand patient evolution in temporal electronic health records, which can be uninformative due to its very general events. We introduce the notion of clusters of hidden states allowing for an expanded understanding of the multiple dynamics that underlie events in such data. Clusters are defined as part of hidden Markov models learned from such data, where the number of hidden states is not known beforehand. We evaluate the proposed approach based on a large dataset from Dutch practices of patients that had events on comorbidities related to atherosclerosis. The discovered clusters are further correlated to medical-oriented outcomes in order to show the usefulness of the proposed method.