Context-Driven Predictions
IJCAI(2007)
摘要
Markov models have been a keystone in Artificial Intelligence for many decades. However, they re- main unsatisfactory when the environment mod- elled is partially observable. There are pathological examples where no history of fixed length is suf- ficient for accurate prediction or decision making. On the other hand, working with a hidden state (like in Hidden Markov Models or Partially Observable Markov Decision Processes) has a high computa- tional cost. In order to circumvent this problem, we suggest the use of a context-based model. Our approach replaces strict transition probabilities by influences on transitions. The method proposed provides a trade-off between a fully and partially observable model. We also discuss the capacity of our framework to model hierarchical knowledge and abstraction. Simple examples are given in or- der to show the advantages of the algorithm.
更多查看译文
关键词
hidden markov models,partially observable markov decision,hidden state,markov model,environment modelled,prediction learning,accurate prediction,artificial intelligence,observable model,fixed length,context-driven prediction,context-based model,associative memories,transition probability,hidden markov model,artificial intelligent,associative memory
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络