Icosystem (United States)

In the field of decision-making a fundamental problem is how to uncover people's choice behavior. While choices them- selves are often observable, our underlying decision strategies determining these choices are not entirely understood. Previous research defined a number of decision strategies and conjectured that people do not apply only one strategy but switch strategies during the decision process. To the best of our knowledge, empirical evidence for the latter conjecture is missing. This is why we monitored the purchase decisions 624 consumers shopping online. We study how many of the observed choices can be explained by the existing strategies in their pure form, how many decisions can be explained if we account for switching behavior, and investigate switching behavior in detail. Since accounting for switching leads to a large search space of possible mixed decision strategies, we apply a genetic algorithm to find the set of mixed decision strategies which best explains the observed behavior. The results show that mixed strategies are used more often than pure ones and that a set of four mixed strategies is able to explain 93.9% of choices in a scenario with 4 alternatives and 75.4% of choices in a scenario with 7 alternatives.

Applying evolutionary algorithms to traditional single-objective optimization problems is now a well-understood and successful process. Less well understood but developing nicely is the use of evolutionary algorithms for multi-objective optimization problems. Least well understood is the use of coevolutionary algorithms to solve co-optimization problems. The dynamics of these coevolutionary processes lead to pathological situations not seen (or easily avoided) in standard optimization algorithms. Such pathologies include endless cycling, loss of fitness gradient, and so on.These problems are due in part to the fact that in many coevolutionary settings even basic notions such as what the goal is or what it means to make progress toward the goal are often vague and ill-defined.Thorough characterizations of possible goals for co-optimization have been introduced and we now have a fairly broad range of so-called ``solution concepts''. Once the idea of a goal was ironed out, the bulk of co-optimization research to date has focused on defining and achieving monotonic progress toward the chosen goal.Via careful analysis and generalizations, this paper reconciles different and apparently conflicting notions of monotonicity existent in the field. However, this analysis also points out that even the weakest monotonicity may simply not be within reach for some solution concepts that are nonetheless of practical interest. Moreover, the monotonicity results tell us nothing about the relative performance of different algorithms.The hypothesis put forward, in light of recent ``free lunch'' results, is that we may be faced with a difficult choice between monotonicity and performance. These ideas are formally developed and analyzed in the context of standard solution concepts such as maximum cumulative utility and best worst case.

This chapter presents a number of illustrative case studies of a wide range of applications of multiobjective optimization methods, in areas ranging from engineering design to medical treatments. The methods used include both conventional mathematical programming and evolutionary optimization, and in one case an integration of the two approaches. Although not a comprehensive review, the case studies provide evidence of the extent of the potential for using classical and modern multiobjective optimization in practice, and opens many opportunities for further research.

In this paper, co-evolution is used to examine the long-term evolution of business models in an industry. Two types of co-evolution are used: synchronous, whereby the entire population of business models is replaced with a new population at each generation, and asynchronous, whereby only one individual is replaced.

This paper focuses on the delay-dependent global stability analysis for a swarm of cooperating networked autonomous vehicles modeled using functional differential equations. The cohesiveness of the swarm under realistic communication constraints such as different communication channel delays and dynamically changing communication topology is the main stability property that we are concerned with. Under these constraints, it cannot be guaranteed that the swarm members will converge to an invariant set of equilibrium points, i.e. the configuration of the swarm members converges to a constant arrangement. The novelty of the result lies in the construction of an appropriate Lyapunov-Krasovskii functional and use of Lyapunov-LaSalle invariance theorem in the context of delay differential equations that when combined produce sufficient conditions which guarantee satisfactory performance of swarm dynamics under communication constraints. An explicit bound on communication delay in dynamically changing communication network topology is obtained to ensure stability of the swarm. These conditions are then used to design stable decentralized motion guidance laws for a swarm of networked vehicles.

In many multiagent systems, small changes in individual-level rules may lead to very large changes at the group-level. This phenomenon is striking in the “aggressor-defender game,” a simple participative game in which each participant randomly selects two others from the group (A and B). In the aggressor game, everyone tries to position themselves so that A is always between themselves and B. In the defender game everyone tries to position themselves between A and B. Despite these exceedingly simple rules and the seemingly small difference between them, the two games exhibit very different dynamics. The aggressor game produces a highly dynamic group that rapidly expands over time whereas the defender game quickly collapses to a tight knot. I analyze these games and provide some insight as to how these two group level behaviors arise, thereby linking the micro- and macro-levels. I also introduce and analyze a new, related and simpler game, the stalker game, in which each participant selects and pursues a single participant, and which also produces a collapsing group. It is suggested that such a geometrical analysis may be applicable for other multiagent systems such as insect societies and collective robotics.

We illustrate with two simple examples how interactive evolutionary computation (IEC) can be applied to exploratory data analysis (EDA). IEC is particularly valuable in an EDA context because the objective function is by definite either unknown a priori or difficult to formalize. The first example IEC is used to evolve the "true" metric of attribute space. Indeed, the assumed distance function in attribute space strongly conditions the information content of a two-dimensional display of the data, regardless of the dimension reduction approach. The goal here is to evolve the attribute space distance function until "interesting" features of the data are revealed when a clustering algorithm is applied. In a second example, we show how a user can interactively evolve an auditory display of cluster data. In this example, we use IEC with genetic programming to evolve a mapping of data to sound functions in order to sonify qualities of data clusters.

As interactive evolutionary computation is starting to penetrate scientific fields that otherwise rely heavily on numerical fitness evaluation, very little emphasis has been put on communicating the relevant information for decision making to the human during the interactive loop. This paper raises the issue of information presentation as a scientific question for interactive evolutionary applications that include automatic fitness evaluation and presents an example for how relevant information display can be for real world problems.

Imitation exerts enormous influence over society, and business and finance in particular. And its influence has grown as the avenues by which people imitate--and are imitated--have multiplied and the process has gotten faster. Thousands of communications channels make it possible for virtually anyone in the developed world to know, almost instantaneously, what others do, think, believe, claim, or predict. More significantly, we can and do act upon such knowledge. The resulting fads and fashions, bubbles and crashes are ever more frequent, severe, and complex. The information age has cast up more than its share of paradoxes, including this one: When information is plentiful, we often use it not to make better decisions but to imitate others--and their mistakes. In consumer purchases, financial markets, and corporate strategy, what others do matters more to us than the facts. When there's too much information, imitation becomes a convenient heuristic. This is the basis for a self-referential society. Imitation has its virtues, but it also promotes instability and unpredictability. That's because, by definition a multiplier, it can swell a single opinion into a mass movement or catapult the smallest player to the forefront of a market. Mastering the dynamics of self-reference won't ensure mastery of its consequences. But businesses that understand how imitation works can at least attempt to gird themselves against its worst effects--by accounting for it in their forecasts and risk-management plans, by becoming more sensitive to unexpectedly changing circumstances, and by avoiding mindless imitation of other companies' moves. In some instances, they may even be able to build strategies around self-reference and use the tools of imitation to capture new business. That won't make the world any less confusing. But it may make it more profitable.

The boards of large corporations sharing some of their directors are connected in complex networks. Boards are responsible for corporations' long-term strategy and are often involved in decisions about a common topic related to the belief in economical growth or recession. We are interested in understanding under which conditions a large majority of boards making the same decision can emerge in the network. We present a model where board directors are engaged in a decision-making dynamics based on "herd behavior." Boards influence each other through shared directors. We find that imitation of colleagues and opinion bias due to the interlock do not trigger an avalanche of identical decisions over the board network, whereas the information about interlocked boards' decisions does. There is no need to invoke global public information, nor external driving forces. This model provides a simple endogenous mechanism to explain the fact that boards of the largest corporations of a country can, in the span of a few months, make the same decisions about general topics.