BAMERS
Inspirado en el proyecto europeo GLODERS que estudia la dinámica de las redes de extorsión, BAMERS es un simulador, implementado en NetLogo, para analizar los efectos de la extorsión en los agregados macro-económicos usuales. La simulación es adecuada para contender con la falta de datos asociada a la naturaleza oculta de la extorsión, que dificulta una aproximación analítica a este problema. A partir de una economía artificial con señales macro-económicas sanas, se introduce la extorsión a diferentes niveles. Los efectos significativos de esta actividad incluyen una relación lineal entre la propensión al consumo y la extorsión, sugiriendo que esta última aumenta la pobreza, con impacto en el índice Gini y la inflación. El producto interno bruto muestra una marcada contracción con la mínima presencia de extorsionadores en el sistema.
BAM
El modelo BAM define una economía compuesta por 3 tipos de agente: empresas, trabajadores y bancos, que interactúan en 3 mercados: laboral, de bienes y crediticio. Las interacciones de estos agentes general la señales macro-económicas de interés: asa de desempleo, producto interno bruto, inflación, distribución de la riqueza, etc. Los agentes están acotados racionalmente, es decir, normalmente deciden con información incompleta. Los mercados definen protocolos fijos de interacción entre los agentes. La señales observadas son propiedades emergentes del sistema. Se trata de una implementación en NetLogo validada y bien documentada del modelo BAM (Delli Gatti et al., 2011).
JaCa-DDM
Download: Github | Sourceforge
JaCa-DDM is a Distributed Data Mining (DDM) tool based on the Agents & Artifacts paradigm as implemented in Jason and CArtAgO. It is intended to design, implement and evaluate learning strategies, i.e., encapsulated workflows modeling the interactions of agents in a distributed environment composed of Weka tools and data sources, with the objective of creating a classification model. JaCa-DDM considers any kind of environment where data is split in various sites, even geographically distributed, as a DDM setting. With JaCa-DDM is possible to configure and deploy a DDM system that takes into account the different sites and their data.
Learning strategies are expected to be applicable in any DDM setting, the deployment details being managed by the JaCa-DDM platform. In this sense, these strategies are plug-and-play. The JaCa-DDM distribution already includes a set of predefined learning strategies, being also possible to add new ones. The actual data mining process is encapsulated on the strategies, which might have some configurable parameters that can be set as part of the general configuration.
JaCa-DDM is a tool for experimenting with different DDM approaches, as it evaluates the produced classification model, yielding various performance statistics (total time, classification accuracy, network traffic produced, model complexity, confusion matrix). It can be extended through the adding of new learning strategies and artifacts. Artifacts are first-class entities in the agent environment that encapsulate services, in the case of JaCa-DDM, these services consists on Weka related tools.
JILDT
Download: Sourceforge.
Jason Induction of Logical Decison Trees (JILDT) is a library that provides an intentional learning mechanism based on induction of logical decision trees to implement learner agents in Jason, the well-known Java-Based implementation of AgentSpeak(L). Top-down Induction of Logical Decision Trees is an Inductive Logic Programming technique, adopted for learning in the context of rational agents. The first-order representation of Tilde is adequate to form training examples as sets of beliefs, e.g., the beliefs of the agent supporting the adoption of a plan as an intention; and the obtained hypothesis are useful for updating the plans and beliefs of the agents, i.e., a Logical Decision Tree expresses hypotheses about the successful or failed executions of the intentions.
Agents defined as instances of this class are able to learn about their reasons to adopt intentions based on their own experience. Two levels in the inductive process have been implemented: A Java-based level with computational performance in mind; and an AgentSpeak(L)-based level which opens the door for some particular forms of social learning. A set of internal actions and plans are provided for allowing the agents to autonomously perform inductive experiments. Implementation details can be reviewed in the published papers.