Tutorial Speaker: (1) Venu Govindaraju, Professor of Computer Science and Engineering, University at Buffalo, State University of New York, USA (2) Anurag Bharadwaj, Senior Research Scientist, eBay Research Labs, USA, (3) Utkarsh Porwal, PhD Student, University at Buffalo, State University of New York,USA
Title: Machine Learning Algorithms and its Applications in Distributed Wireless Sensor Networks
Wireless Sensor Networks (WSNs) has been applied in various domains such as environment monitoring, industry and healthcare for various purposes. WSNs are used to interact with the environment and send useful information across the network. These sensors need to keep track of states of other nodes, perform adaptive information extraction, and predict location of mobile nodes for optimum utilization of resources. However, sensors often suffer from problems such as physical constraints, low bandwidth and ineffective communication between different nodes which leads to error and inefficiency in the network. Hence, WSNs have inspired research on machine learning techniques to address these limitations with more generic ways of extracting information and discovering relations. Machine learning has already been used to address problems such as to make
inference from the data by extracting useful features, building similarity matrices, mapping vector spaces of signals given by sensors with physical spaces and dealing with large amount of online unlabeled data. Different learning paradigms such as supervised, unsupervised and reinforcement learning has been used to address these issues. Moreover, classification, regression and kernel techniques have been used to solve several problems pertaining to WSNs. On the other hand Machine learning and Artificial Intelligence community has taken great interest in distributed wireless sensor networks. Applications such as robotics and context aware computing have used WSNs extensively for more intelligent and automated learning process. This tutorial intends to give the comprehensive overview of fundamental working and limitations of such machine learning algorithms and highlight the scope of future research in this area using Machine Learning.