Intelligent instance selection of data streams for smart sensor applications

Abstract

The purpose of our work is to mine streaming data from a variety of hundreds of automotive sensors in order to develop methods to minimize driver distraction from in-vehicle communications and entertainment systems such as audio/video devices, cellphones, PDAs, Fax, eMail, and other messaging devices. Our endeavor is to create a safer driving environment, by providing assistance in the form of warning, delaying, or re-routing, incoming signals if the assistance system detects that the driver is performing, or is about to perform, a critical maneuver, such as passing, changing lanes, making a turn, or during a sudden evasive maneuver. To accomplish this, our assistance system relies on maneuver detection by continuously evaluating various embedded vehicle sensors, such as speed, steering, acceleration, lane distance, and many others, combined into representing an instance of the "state" of the vehicle. One key issue is how to effectively and efficiently monitor many sensors with constant data streams. Data streams have their unique characteristics and may produce data that is not relevant or pertinent to a maneuver. We propose an adaptive sampling method that takes advantage of these unique characteristics and develop algorithms that attempt to select relevant and important instances to determine which sensors to monitor and how to provide quick and effective responses to this type of mission critical situations. This work can be extended to many similar sensor applications with data streams.

Context-aware adaptive data stream mining

“…A clustering technique has been used for analyzing and detecting the driver behavior. In [12], real-time mining of information from in-vehicle sensors to minimize driver distraction is proposed through adaptation of the instance-selection process based on changes to the data distribution. While this research recognizes the implicit need for adaptation, it is focused on intelligent sampling with little consideration for resource availability.…”

“…The review of the current state-of-the-art of data stream mining in resource-constrained environments indicates that there are data stream mining algorithms [12,15,16] that are able to function on devices such as PDAs but they have limited ability to cope with a multitude of changing contextual factors. A context-aware adaptation approach leverages the full potential of UDM and can provide continuity and consistency of the running application by adapting parameters of mining algorithms according to context changes.…”

Interactive self-adaptive clutter-aware visualisation for mobile data mining

“…In the last few years, rapid strides have been made in accurately and efficiently mining high speed data streams [15] in mobile devices such as Personal Digital Assistants (PDAs) [10,11,6] and there is a growing focus on "in-network" processing using embedded devices such as sensor nodes [7,12,13]. Kargupta et al [11] have developed a client/server data stream mining system: MobiMine which focuses on data stream mining applications for stock market data.…”

Situation-Aware Adaptive Processing (SAAP) of Data Streams

“…Here we consider context as the information used for representing real-world situations [4] in pervasive computing environments. Con-textual information collected from every single sensor or data source represents a partial view of the real-world.…”

“…Reviewing recent works in mobile data stream mining reveals that most of these projects [1][2][3][4] have limited levels of adaptations or mainly focusing on the battery or memory usage. A general approach for smart and cost-efficient analysis of data that is under-pinned using situation-aware adaptation has not been introduced in the current state-of-the-art and is still an open issue.…”