Multi-task Learning for Software Agents
نویسنده
چکیده
This paper describes experiments done to demonstrate the effectiveness of Multi-task Learning (MTL) for software agents. The experiments were carried out on an agent for processing electronic mail (email). MTL is shown to slightly improve the learning rate in this domain over Single-task Learning (STL) in a k-nearest-neighbor implementation. We then introduce a new method of lazy learning we call Neural Neighbor which lends itself much better to the incorporation of MTL and outperforms both our STL and MTL k-nearest-neighbor implementations.
منابع مشابه
Improving Agent Performance for Multi-Resource Negotiation Using Learning Automata and Case-Based Reasoning
In electronic commerce markets, agents often should acquire multiple resources to fulfil a high-level task. In order to attain such resources they need to compete with each other. In multi-agent environments, in which competition is involved, negotiation would be an interaction between agents in order to reach an agreement on resource allocation and to be coordinated with each other. In recent ...
متن کاملTask Similarity-Based Task Allocation Approach in Multi-Agent Engineering Software Systems
Current complex engineering software systems are often made up of many software components to perform complex tasks, which can be modeled as multi-agent systems. Task allocation in complex multi-agent engineering software systems can be described through software agents’ cooperation to satisfy the resource requirement of tasks. Although many task allocation approaches have been presented to dea...
متن کاملDynamic Agent Communities Facilitating to Distant Learning in a Virtual University Information Space
-Presented is the approach to the design of Virtual Information Space for a virtual/real university based upon the paradigms of intellectual software agent, multi-agent system, dynamic task-oriented agent community. Virtual Information Space is proposed as the unifying concept for the design of the distributed intelligent distant learning facilities. The particularity of the framework presented...
متن کاملLooking for Scalable gents
Reinforcement Learning intends to ease and possibly to perform automatically the design of systems such as software or robot agents. An important aspect is the ability of learning agents to adapt to their environment and to the task they have to accomplish. This kind of learning is unfortunately restrained by problems like combinatorial explosion of the state space that limits the number of sen...
متن کاملCollective Machine Learning: Team Learning and Classification in Multi-Agent Systems
This dissertation focuses on the collaboration of multiple heterogeneous, intelligent agents (hardware or software) which collaborate to learn a task and are capable of sharing knowledge. The concept of collaborative learning in multi-agent and multi-robot systems is largely under studied, and represents an area where further research is needed to gain a deeper understanding of team learning. T...
متن کامل