In order to identify the bacteria in wound infection, an electronic nose system with a sensor array of 34 sensors was designed. Eight kinds samples were detected, i.e., culture medium, Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa and their mixture different concentration. Using support vector machine as classifier without optimization, recognition rate is up 86.54%. To simplify improve for samples, Wilks' lambda statistic (Wilks' Λ-statistic), Mahalanobis distance,...

Security is one of the major concerns in Industrial Wireless Sensor Networks (IWSNs). To assure security clustered IWSNs, this paper presents a secure clustering protocol with fuzzy trust evaluation and outlier detection (SCFTO). Firstly, to deal transmission uncertainty an open wireless medium, interval type-2 logic controller adopted estimate trusts. And then density based mechanism introduced acquire adaptive threshold used isolate malicious nodes from being cluster heads. Finally, heads...

A typical wireless sensor network (WSN) is often deployed in harsh or hostile environment to work an unattended mode. It prone various attacks during routing process as the open medium. Traditional encryption and authentication-based secure mechanisms cannot avoid from internal compromised nodes, they also need much resource for complex computing. To conquer these problems, trust-based mechanism usually used WSNs. However, how get behavior evidences trust assessment rarely addressed date. In...

Diabetic peripheral neuropathy (DPN) is an early developing complication of diabetes mellitus associated with nerve dysfunction. Artesunate (ART), a natural compound extracted from the herb Artemisia annua L., was reported to benefit neural injury. However, whether ART has role in preventing DPN still unknown. In this study, rat model high fat diet feeding and streptozotocin (STZ) injection established. The findings indicated that treatment significantly ameliorated hyperglycemia-induced hot...

Trustworthy and reliable data delivery is a challenging task in Wireless Sensor Networks (WSNs) due to unique characteristics constraints. To acquire secured address the conflict between security energy, this paper we present an evolutionary game based secure clustering protocol with fuzzy trust evaluation outlier detection for WSNs. Firstly, method presented transform transmission evidences into values while effectively alleviating uncertainty. And then, K-Means scheme proposed further...

In recent years, the electronic nose (E-nose) technology has made a great progress in practical application. However, interferences, such as interferences from non-target gas, manufacturing deviation and sensor drift, have negative impact on detection accuracy during real-time process, which will affect industrialization commercialization of E-nose. Thus, it is necessary to research method interference suppression for realizing long-term stable detection. This paper proposes sparse...

The feature extraction technique for an electronic nose (e-nose) applied in tobacco smell detection open country/outdoor environment with periodic background strong interference is studied this paper. Principal component analysis (PCA), Independent (ICA), re-filtering and a priori knowledge are combined to separate suppress on the e-nose. By coefficient of multiple correlation (CMC), it can be verified that better separation environmental temperature, humidity, atmospheric pressure variation...

This paper gives a solution to solve the interference problem of electronic nose (e-nose), which is ill-posed due uncertainty and unpredictability its instable behavior. Traditional methods for suppression are component correction frameworks, laborious or little efficient. With (especially background sensor drift), distribution test data obtained in practical application different from that previous training data. From viewpoint machine learning, novel domain adaptive extreme learning...

Butylparaben (BuP), as an emerging contaminant with endocrine-disrupting effects, may exert effects on skin pigmentation in fish by interfering the neuroendocrine system. Therefore, models of BuP exposure Nile tilapia (Oreochromis niloticus) were established adding different doses (0, 5, 50, 500, and 5000 ng/L) for 56 days. The obtained results showed that induced darker pigmentation, manifested increased melanin content skin, while genes related to synthesis, including α-MSH Asip2,...

This paper proposes a way for drift compensation in electronic noses (e-nose) that often suffers from uncertain and unpredictable sensor drift. Traditional machine learning methods odor recognition require consistent data distribution, which makes the model trained with previous less generalized. In actual application scenario, collected previously later may have different distributions due to If dataset without is treated as source domain target domain, correction based on kernel...

Electronic nose (E-nose), a kind of instrument which combines with the gas sensor and corresponding pattern recognition algorithm, is used to detect type concentration gases. However, drift will occur in realistic application scenario E-nose, makes variation data distribution feature space causes decrease prediction accuracy. Therefore, studies on compensation algorithms are receiving increasing attention field E-nose. In this paper, novel method, namely Wasserstein Distance Learned Feature...

Summary The main purpose of the present work is to predict permeability a porous medium from its three-dimensional (3D) structure network. In this work, 3D reconstructed by truncated Gaussian method using Fourier transform and starting 2D binary image obtained thin section sample. skeleton provides way visualizing graph pore It determined thinning algorithm, which conceived preserve topology. gives both visual quantitative information about connectivity space, coordination number for every...

A medical electronic nose (e-nose) with 31 gas sensors is used for wound infection detection by analyzing the bacterial metabolites. In practical applications, prediction accuracy drops dramatically when model established laboratory data directly in human clinical samples. This a key issue e-nose which should be more worthy of attention. The host (carrier) bacteria can culture solution, animal wound, or wound. As well, solution animals (such as: mice, rabbits, etc.) obtained easily are...

An intelligent and automatic control method for tobacco flue-curing based on machine learning was proposed which makes use both of a priori rules (human knowledge) curing collected feature data about during the period.