Data Stream Mining using Reinforcement Learning: A Survey

Apply

Data Stream Mining using Reinforcement Learning: A Survey

Journal Title: International Journal of Research in Computer and Communication Technology - Year 2014, Vol 3, Issue 1

Abstract

Large numbers of real world applications are producing continuous and large data streams for further analysis. Data stream mining is allowing users to extracts knowledge in the form of patterns or model from the continuous data streams of information. With limitations of data stream mining, it has many challenges in storage, computation. Also the data streams are subjected to changes with environment and time. So while designing the data stream mining process it is important to have flexibility to promptly adapt with changing concepts and data distribution depending on nature of data. The Reinforcement learning is an approach to learning by interacting with the environment. This method builds better tree over time as it learns from the feedback provided by the previous building operations. The objective of this article is to analyze the various classification applications used as one of data mining techniques. Also we reviewed reinforcement learning framework, methods of reinforcement learning, and building a decision tree with reinforcement learning for time changing data.

Authors and Affiliations

Pramod Patil, Ashwini Ahirrao

Keywords

EP ID EP27822
DOI -
Views 279
Downloads 0