Advance in Electronic and Electric Engineering. ISSN 2231-1297, Volume 4, Number 5 (2014), pp. 475-486 © Research India Publications http://www.ripublication.com/aeee.htm Machine Learning for Wireless Sensor Network: A Review, Challenges and Applications Mayur V. Bhanderi 1 and Hitesh B. Shah 2 1 Department of Electronics and Communication, A D Patel Institute of Technology, New Vallabh Vidhyanagar Nagar, Gujarat, India. 2 Department of Electronics and communication, G H Patel College of Engineering and Technology, Vallabh Vidhyanagar Nagar, Gujarat, India. Abstract In recent years, there has been a growing attention in wireless sensor networks. In Wireless Sensor Networks (WSNs), collecting sensed information, transforming the information data to the base station in an energy efficient way, and lengthening the network lifetime are main issues. Sensor nodes in WSNs are energy constrained. So, one of the major design challenges in WSNs is minimizing consumed energy at the sensor nodes. Therefore, a number of routing schemes are designed that make efficient usage of limited energy of the sensor nodes. Hierarchical routing protocols are best known in regard to energy efficiency. By using a clustering technique hierarchical routing protocols greatly reduce energy consumed in collecting and disseminating data. But, large-scale sensor network unavoidably introduce large amount of data in WSNs to be processed, transmitted and received. Transmitting all data back to a base station for processing and making inferences is merely impossible due to the sensor limited energy and bandwidth constraints. Thus, there is a need for applying Machine Learning (ML) algorithms in WSNs. These algorithms could significantly decrease the amount of data communications and truthfully utilize the distributive characteristic of WSNs. The main goal of this review paper is to demonstrate that ML is a practical approach to a range of complex distributed problems in WSNs and particularly in energy efficient routing.