Thomas Vallée

Off-Line Computation of Stackelberg Solutions with the Genetic Algorithm

The paper studies off-line computation of the Stackelberg solution in a repeated game framework, utilizing the Genetic Algorithm. Simulations are conducted with a numerical linear quadratic example and a Fish War game example. Furthermore, it is shown that an evolutionary mutation probability is preferable to a fixed one as usually assumed.

Social and technological efficiency of patent systems

This article develops an evolutionary model of industry dynamics in order to carry out a richer theoretical analysis of the consequences of a stronger patent system. The first results obtained in our article are rather consistent with the anti-patent arguments and they do not favour the case for a stronger patent system: higher social welfare and technical progress are observed in our model in industries with milder patent systems (lower patent height and patent life).

Optimal open loop cheating in dynamic reversedLinear ‐ Quadratic Stackelberg games

The distinctive characteristic of a “Reversed Stackelberg Game” is that the leader playstwice, first by announcing his future action, second by implementing a possibly differentaction given the followers reaction to his announcement. In such a game, if the leader usesthe normal Stackelberg solution to find (and announce) his optimal strategy, there is a strongtemptation for him to cheat, that is, to implement another action than the one announced. Inthis paper, within the framework of a standard discrete time Linear‐Quadratic DynamicReversed Stackelberg game, we discuss and derive the best possible open‐loop cheatingstrategy for an unscrupulous leader.

Range-based Multi-Actor Multi-Criteria Analysis: A combined method of Multi-Actor Multi-Criteria Analysis and Monte Carlo simulation to support participatory decision making under uncertainty

Abstract Concerns about environmental and social effects have made Multi-Criteria Decision Making (MCDM) increasingly popular. Decision making in complex contexts often – possibly always – requires addressing an aggregation of multiple issues to meet social, economic, legal, technical, and environmental objectives. These values at stake may affect different stakeholders through distributional effects characterized by a high and heterogeneous uncertainty that no social actors can completely control or understand. On this basis, we present a new process framework that aims to support participatory decision making under uncertainty: the range-based Multi-Actor Multi-Criteria Analysis (range-based MAMCA). On the one hand, the process framework explicitly considers stakeholders’ objectives at an output level of aggregation. On the other hand, by means of a Monte Carlo analysis, the method also provides an exploratory scenario approach that enables the capture of the uncertainty, which stems from the complex context evolution. Range-based MAMCA offers a unique participatory process framework that enables us (1) to identify the alternatives pros and cons for each stakeholder group; (2) to provide probabilities about the risk of supporting mistaken, or at least ill-suited, decisions because of the uncertainty regarding to the decision-making context; (3) to take the decision-makers’ limited control of the actual policy effects over the implementation of one or several options into account. The range-based MAMCA framework is illustrated by means of our first case study that aimed to assess French stakeholders’ support for different biofuel options by 2030.

Control variation as a source of uncertainty: Single input case

This paper presents a theoretical framework and the control strategy for single input discrete-time stochastic systems for which the control variations increase state uncertainty (CVIU systems). This type of system model can be useful in many practical situations, such as in monetary policy problems, medicine and biology, and, in general, in problems for which a complete dynamic model is too complex to be feasible. The optimal control strategy for a single-input CVIU system associated with a convex cost functional is devised using dynamic programming and tools from nonsmooth analysis. Furthermore, this strategy points to a region in the state space in which the optimal action is of no variation, as expected from the cautionary nature of controlling underdetermined systems. In addition, a specific result for the case when the cost functional is differentiable is obtained and discussed. These results are illustrated through a numerical example in economics.

Computation of the Closed-Loop Stackelberg Solution Using the Genetic Algorithm

Abstract This paper deals with the computation of the closed loop dynamic Stackel-berg equilibrium solution in two-player nonzero-sum dynamic games using the genetic algorithm. When the leader has access to closed-loop state information, which provides him (indirectly) with on-line information on the past actions of the follower, derivation of the Stackelberg solution is known to be a challenging one. We address here the question of whether in this context genetic algorithm techniques or their appropriately modified versions can be used as computational tools. Following demonstrations using analytic derivations in general linear quadratic games, the paper applies the tool to a nonlinear dynamic taxation problem.

Learning how to regulate a polluter with unknown characteristics: An application of genetic algorithms to a game of dynamic pollution control

We consider, within the framework of a dynamic game, the problem of a regulator using taxes to force a polluting monopolist to act in a socially optimal way. Traditionally it has been assumed that, in such a case, the regulator will implement his Stackelberg solution. However, the Stackelberg solution presupposes that the regulator knows exactly all characteristics of the monopolist. We show here that the on-line use of a genetic algorithm may allow the regulator to approximate the Stackelberg solution, even when he has only extremely limited information about the monopolist. While the results presented here are still preliminary, the fast convergence towards the analytic solution appears to suggest that the approach may be of practical value in real situations.

Similarity in trade structure: An evidence from ASEAN + 3

ABSTRACT The present paper aims to explore the competition in exports among ASEAN + 3 members by applying the export similarity index for the disaggregation export data from 1990 to 2014. We also discuss the changes of comparative advantage of ASEAN + 3s exports through the revealed comparative advantage index and its relation with similarity index. First, the results support the idea that export similarity varies among ASEAN + 3 member states over the period 1990–2014. Second, we find evidence of several export communities. In particular, five countries, notably China, Japan, Malaysia, Singapore and South Korea, with the highest similarity index form an export community mainly in the electronic microcircuits and parts of office machines. Third, the revealed comparative advantage seems to be the key factor defining the similarity level of exports. Finally, the revealed comparative advantage analysis partly allows us to verify the Ricardos theory and New Trade theory in the context of ASEAN + 3 trade integration.