Towards an Online Seizure Advisory System-An Adaptive Seizure Prediction Framework Using Active Learning Heuristics.

In the last decade, seizure prediction systems have gained a lot of attention because of their enormous potential to largely improve the quality-of-life of the epileptic patients. The accuracy of the prediction algorithms to detect seizure in real-world applications is largely limited because the brain signals are inherently uncertain and affected by various factors, such as environment, age, drug intake, etc., in addition to the internal artefacts that occur during the process of recording the brain signals. To deal with such ambiguity, researchers transitionally use active learning, which selects the ambiguous data to be annotated by an expert and updates the classification model dynamically. However, selecting the particular data from a pool of large ambiguous datasets to be labelled by an expert is still a challenging problem. In this paper, we propose an active learning-based prediction framework that aims to improve the accuracy of the prediction with a minimum number of labelled data. The core technique of our framework is employing the Bernoulli-Gaussian Mixture model (BGMM) to determine the feature samples that have the most ambiguity to be annotated by an expert. By doing so, our approach facilitates expert intervention as well as increasing medical reliability. We evaluate seven different classifiers in terms of the classification time and memory required. An active learning framework built on top of the best performing classifier is evaluated in terms of required annotation effort to achieve a high level of prediction accuracy. The results show that our approach can achieve the same accuracy as a Support Vector Machine (SVM) classifier using only 20 % of the labelled data and also improve the prediction accuracy even under the noisy condition.

Symbiotic Sensing for Energy-Intensive Tasks in Large-Scale Mobile Sensing Applications.

Energy consumption is a critical performance and user experience metric when developing mobile sensing applications, especially with the significantly growing number of sensing applications in recent years. As proposed a decade ago when mobile applications were still not popular and most mobile operating systems were single-tasking, conventional sensing paradigms such as opportunistic sensing and participatory sensing do not explore the relationship among concurrent applications for energy-intensive tasks. In this paper, inspired by social relationships among living creatures in nature, we propose a symbiotic sensing paradigm that can conserve energy, while maintaining equivalent performance to existing paradigms. The key idea is that sensing applications should cooperatively perform common tasks to avoid acquiring the same resources multiple times. By doing so, this sensing paradigm executes sensing tasks with very little extra resource consumption and, consequently, extends battery life. To evaluate and compare the symbiotic sensing paradigm with the existing ones, we develop mathematical models in terms of the completion probability and estimated energy consumption. The quantitative evaluation results using various parameters obtained from real datasets indicate that symbiotic sensing performs better than opportunistic sensing and participatory sensing in large-scale sensing applications, such as road condition monitoring, air pollution monitoring, and city noise monitoring.

Admicellar polymerization and characterization of thin poly(2,2,2-trifluoroethyl acrylate) film on aluminum alloys for in-crevice corrosion control.

Corrosion control of aluminum alloys in the aerospace industry has been of great interest in recent years, especially the aging of certain fleets in the United States Air Force. A thin film of poly(2,2,2-trifluoroethyl acrylate) (PTFEA) has been deposited on aluminum alloy coupons by admicellar polymerization for the purpose of in situ control of corrosion in narrow gaps. Polymerization conditions were chosen based on contact angle measurements, and the final product film was characterized using Fourier transform infrared spectroscopy, scanning electron microscopy, atomic force microscopy, and X-ray photoelectron spectroscopy. Surface characterization studies have shown that the polymeric film is approximately 10 nm thick with nonuniform deposition at this scale. The modified surface is highly hydrophobic and able to delay salt solution uptake (3.5 wt % NaCl) for a period of up to 6 h in crevice corrosion tests. PTFEA films reduced the corroded area to 20% compared to 65% for a bare aluminum control and to 33% for poly(methyl methacrylate) (PMMA) film in a 24 h crevice test. PTFEA film exhibits better corrosion protection than PMMA film because it has higher hydrophobicity than a PMMA-modified surface and comparable properties as a corrosion barrier.