Multiagent Environment for Hybrid AI Models
نویسندگان
چکیده
To utilize the the concept of combining various modern artificial intelligence methods — namely neural networks, genetic algorithms and fuzzy logic controllers — we develop a unified software platform called Bang2. The system serves as a library for many artificial intelligence methods and allows for easy creation of such combinations, and possibly their semi-automated generation, or even evolution. It also ensures their deployment through the computer network and parallel processing. Inspired by software agents paradigm we have designed Bang2 as a community of cooperating autonomous software agents. Key-Words: Hybrid methods, Multiagent systems, Adaptive agents.
منابع مشابه
Solving the flexible job shop problem by hybrid metaheuristics-based multiagent model
The flexible job shop scheduling problem (FJSP) is a generalization of the classical job shop scheduling problem that allows to process operations on one machine out of a set of alternative machines. The FJSP is an NP-hard problem consisting of two sub-problems, which are the assignment and the scheduling problems. In this paper, we propose how to solve the FJSP by hybrid metaheuristics-based c...
متن کاملMultiagent Systems: Challenges and Opportunities for Decision-Theoretic Planning
the integration of two distinct lines of AI research: (1) decision-theoretic planning (DTP) and (2) multiagent systems. Both areas (especially the second) are attracting considerable interest, but work in multiagent systems often assumes either classical planning models or prespecified economic valuations on the part of the agents in question. By integrating models of DTP in multiagent systems ...
متن کاملHalf Field Offense: An Environment for Multiagent Learning and Ad Hoc Teamwork
The RoboCup 2D simulation domain has served as a platform for research in AI, machine learning, and multiagent systems for more than two decades. However, for the researcher looking to quickly prototype and evaluate different algorithms, the full RoboCup task presents a cumbersome prospect, as it can take several weeks to set up the desired testing environment. The complexity owes in part to th...
متن کاملStudent-Friendly Java-Based Multiagent Event Handling
This work describes a student-friendly, pedagogy-oriented event-handling system for managing multiagent AI simulations in a classroom environment.
متن کاملA Hybrid Optimization Algorithm for Learning Deep Models
Deep learning is one of the subsets of machine learning that is widely used in Artificial Intelligence (AI) field such as natural language processing and machine vision. The learning algorithms require optimization in multiple aspects. Generally, model-based inferences need to solve an optimized problem. In deep learning, the most important problem that can be solved by optimization is neural n...
متن کامل