Uninterrupted Learning Agent Using Finite State
نویسندگان
چکیده
In a multi-agent system (MAS), an issue arises as agent may possibly encounter a new event, an event that has never been perceived by the agent or was never anticipated by a human developer during the agent implementation. The ability to recognise an event as being on a new class, without having to shut the system down, would significantly enhance the agent's learning potentials. This paper presents the means for allowing an agent to recognise new events without having to interrupt its services. Our scheme is based on a dynamic construction of finite-state machines (FSM) from the general regularities implicitly embedded in the data. We have extended additional functionality in FSM transition function to enhance the recognition. Our experiment has shown that our extended FSM can enhance the learning process and allows the system to continue working without system interruption.
منابع مشابه
Utilizing Generalized Learning Automata for Finding Optimal Policies in MMDPs
Multi agent Markov decision processes (MMDPs), as the generalization of Markov decision processes to the multi agent case, have long been used for modeling multi agent system and are used as a suitable framework for Multi agent Reinforcement Learning. In this paper, a generalized learning automata based algorithm for finding optimal policies in MMDP is proposed. In the proposed algorithm, MMDP ...
متن کاملLearning and Planning in Multiagent POMDPs Using Finite-State Models of Other Agents
My thesis work provides a new framework for planning in multiagent, stochastic, partially observable domains with little knowledge about other agents. The relevance of the contribution lays in the variety of practical applications this approach can help tackling, given the very generic assumptions about the environment and the other agents. In order to cope with this level of generality, Bayesi...
متن کاملIncremental PDFA Learning for Conversational Agents
When finite-state machines are used for dialogue models of a conversational agent, learning algorithms which learn probabilistic finite-state automata with the state merging method are useful. However, these algorithms should learn the whole data every time the number of example dialogues increases. Therefore, the learning cost is large when we construct dialogue models gradually. We proposed a...
متن کاملNonparametric Bayesian Learning of Other Agents? Policies in Interactive POMDPs
We consider an autonomous agent facing a partially observable, stochastic, multiagent environment where the unknown policies of other agents are represented as finite state controllers (FSCs). We show how an agent can (i) learn the FSCs of the other agents, and (ii) exploit these models during interactions. To separate the issues of off-line versus on-line learning we consider here an off-line ...
متن کاملHierarchical Learning from Demonstration on Humanoid Robots
Developing behaviors for humanoid robots is difficult due to the high complexity of programming these robots, which includes repeated trial and error cycles. We have recently developed a learning from demonstration system capable of training agent behaviors from a small number of training examples. Our system represents a complex behavior as a hierarchical finite automaton, permitting decomposi...
متن کامل