Search

The handheld diagnosis and analysis are highly dependent on the physiological data in the clinical sector. Detection of the defect in the neuronal-assisted activity raises the challenge to the prevailing treatment that benefits from machine learning approaches. The congregated EEG data is then utilized in design of learning applications to develop a model that classifies intricate EEG patterns into active and inactive segments. During arithmetic problem-solving EEG signal acquired from frontal lobe contributes for intelligence detection. The low intricate statistical parameters help in understanding the objective. The mean of the segmented samples and standard deviation are the features extracted for model building. The feature selection is handled using correlation and Fisher score between {Fp1 and F8} and priority ranking of the regions with enhanced activity are selected for the classifier models to the training net. The R-studio platform is used to classify the data based on active and inactive liability. The radial basis function kernel for support vector machine (SVM) is deployed to substantiate the proposed methodology. The vulnerable regions F1 and F8 for arithmetic activity can be visualized from the correlation fit performed between regions. Using SVM classifier sensitivity of 92.5% is obtained for the selected features. A wide range of clinical problems can be diagnosed using this model and used for brain-computer interface. [ABSTRACT FROM AUTHOR]

Because of their thermophysical and rheological properties, graphene oxide (GO) nanofluids show promising advances in heat transfer enhancement. In particular, in magnetohydrodynamic (MHD) studies, where the fluid flow is kept in check, heat transfer tends to diminish due to magnetic field strength. GO nanoparticles, with the highest thermal conductivity, significantly impacts heat transfer devices through conductive heat transfer enhancement. This paper computationally investigates the MHD flow of GO nanofluid over a linearly stretching cylinder. The nanofluid flow is modelled using Buongiorno model under the influence of viscous dissipation effects and the effects of nanoparticle characteristics such as thermophoresis and Brownian motion. The modelled equations are solved using spectral collocation method under isothermal and slip boundary conditions. An examination of the impacts of embedded parameters is presented in detail and it is shown that the conductive heat transfer and diffusive mass transfer are enhanced by dispersing GO nanoparticles in the base fluid. A quantitative analysis is made with the previously published results for special cases. As suggested, this study is significant in heat transfer applications which demand the use of magnetic fields. [ABSTRACT FROM AUTHOR]