Showing papers by "Chung Yuan Christian University published in 2017"

SVM and SVM Ensembles in Breast Cancer Prediction.

Abstract: Breast cancer is an all too common disease in women, making how to effectively predict it an active research problem. A number of statistical and machine learning techniques have been employed to develop various breast cancer prediction models. Among them, support vector machines (SVM) have been shown to outperform many related techniques. To construct the SVM classifier, it is first necessary to decide the kernel function, and different kernel functions can result in different prediction performance. However, there have been very few studies focused on examining the prediction performances of SVM based on different kernel functions. Moreover, it is unknown whether SVM classifier ensembles which have been proposed to improve the performance of single classifiers can outperform single SVM classifiers in terms of breast cancer prediction. Therefore, the aim of this paper is to fully assess the prediction performance of SVM and SVM ensembles over small and large scale breast cancer datasets. The classification accuracy, ROC, F-measure, and computational times of training SVM and SVM ensembles are compared. The experimental results show that linear kernel based SVM ensembles based on the bagging method and RBF kernel based SVM ensembles with the boosting method can be the better choices for a small scale dataset, where feature selection should be performed in the data pre-processing stage. For a large scale dataset, RBF kernel based SVM ensembles based on boosting perform better than the other classifiers.

Major depression detection from EEG signals using kernel eigen-filter-bank common spatial patterns

Abstract: Major depressive disorder (MDD) has become a leading contributor to the global burden of disease; however, there are currently no reliable biological markers or physiological measurements for efficiently and effectively dissecting the heterogeneity of MDD. Here we propose a novel method based on scalp electroencephalography (EEG) signals and a robust spectral-spatial EEG feature extractor called kernel eigen-filter-bank common spatial pattern (KEFB-CSP). The KEFB-CSP first filters the multi-channel raw EEG signals into a set of frequency sub-bands covering the range from theta to gamma bands, then spatially transforms the EEG signals of each sub-band from the original sensor space to a new space where the new signals (i.e., CSPs) are optimal for the classification between MDD and healthy controls, and finally applies the kernel principal component analysis (kernel PCA) to transform the vector containing the CSPs from all frequency sub-bands to a lower-dimensional feature vector called KEFB-CSP. Twelve patients with MDD and twelve healthy controls participated in this study, and from each participant we collected 54 resting-state EEGs of 6 s length (5 min and 24 s in total). Our results show that the proposed KEFB-CSP outperforms other EEG features including the powers of EEG frequency bands, and fractal dimension, which had been widely applied in previous EEG-based depression detection studies. The results also reveal that the 8 electrodes from the temporal areas gave higher accuracies than other scalp areas. The KEFB-CSP was able to achieve an average EEG classification accuracy of 81.23% in single-trial analysis when only the 8-electrode EEGs of the temporal area and a support vector machine (SVM) classifier were used. We also designed a voting-based leave-one-participant-out procedure to test the participant-independent individual classification accuracy. The voting-based results show that the mean classification accuracy of about 80% can be achieved by the KEFP-CSP feature and the SVM classifier with only several trials, and this level of accuracy seems to become stable as more trials (i.e., <7 trials) are used. These findings therefore suggest that the proposed method has a great potential for developing an efficient (required only a few 6-s EEG signals from the 8 electrodes over the temporal) and effective (~80% classification accuracy) EEG-based brain-computer interface (BCI) system which may, in the future, help psychiatrists provide individualized and effective treatments for MDD patients.

Flame Images for Oxygen Content Prediction of Combustion Systems Using DBN

Abstract: As an increasingly popular method in the machine learning field, deep learning is applied to industrial combustion processes in this work. Using easily available color flame images obtained by the charge-coupled device (CCD), a soft sensor system based on deep learning is proposed to predict the outlet oxygen content online. Unlike the traditional principal component analysis which only extracts linear features, a multilayer deep belief network (DBN) is designed to extract the nonlinear features for a better description of the important trends in a combustion process. With the DBN-based multilevel representation of the CCD flame images, more useful information about the physical properties of a flame can be characterized. Sequentially, in a supervised fine-tuning stage, two DBN-based regression models are simply constructed to obtain the nonlinear relationship between the flame images and the outlet oxygen content. The advantages of the proposed deep learning-based analyzing and modeling method are demons...

Insight into the adsorption mechanism of cationic dye onto biosorbents derived from agricultural wastes

Abstract: This study investigated the phenomenon and mechanism of adsorption of methylene green 5 (MG5) on three pristine biosorbents: golden shower pod (GS), coconut shell (CC), and orange peel (OP). The results showed that the biosorbents possessed low specific surface areas, but abundant functional groups. Adsorption was strongly affected by the solution’s pH and ionic strength. As revealed in the kinetic study, equilibrium was rapidly established, requiring low activation energies; a removal rate of 30%–87% was achieved within 1 min. The maximum Langmuir adsorption capacities at 30°C exhibited the following order: GS (106 mg/g) > OP (92 mg/g) > CC (59 mg/g). Thermodynamic experiments suggested that the adsorption occurred spontaneously (−ΔG°) and exothermically (−ΔH°). The primary adsorption mechanisms involved electrostatic attraction, hydrogen bonding formations, and n-π interaction. Thermogravimetric analysis (TGA) revealed that three biopolymer components (i.e., hemicellulose, cellulose, and lignin)...

Eu3+-Doped Phosphor-in-Glass: A Route toward Tunable Multicolor Materials for Near-UV High-Power Warm-White LEDs

Abstract: A novel route toward tunable multicolor materials through phosphor-in-glass (PiG) technology is proposed in this work. Before that, an ultrastable Eu3+-doped precursor luminescent glass frit without thermal quenching in the temperature range of 80–480 K is developed to serve as an encapsulant that not only protects the embedded phosphor but also provides the red-emitting component for the PiG. By adjusting the mass ratio of Sr4Al14O25:Eu2+ phosphor to glass frit, a series of tunable multicolor Eu3+-doped PiG is obtained and exhibits a good resistance to the harsh conditions. Meanwhile, the luminescent color of Eu3+-doped PiG can be modified by changing the excitation wavelength or ambient temperature. Finally, corresponding Eu3+-doped PiG encapsulated high-power light-emitting diodes are further fabricated, especially the warm white-light-emitting diodes (WLEDs), showing good color stability under different drive currents and with different periods of operating time. All these results indicate that Eu3+-doped PiG is a potential color converter applied in the high-power warm-WLEDs and the route above opens up a facile and potential approach to obtain multicolor materials.

Electricity consumption prediction using a neural-network-based grey forecasting approach

Abstract: Electricity consumption is an important economic index and plays a significant role in drawing up an energy development policy for each country. Multivariate techniques and time-series analysis have been proposed to deal with electricity consumption forecasting, but a large amount of historical data is required to obtain accurate predictions. The grey forecasting model attracted researchers by its ability to characterize an uncertain system effectively with a limited number of samples. GM(1,1) is the most frequently used grey forecasting model, but its developing coefficient and control variable were dependent on the background value that is not easy to be determined, whereas a neural-network-based GM(1,1) model called NNGM(1,1) has been presented to resolve this troublesome problem. This study has applied NNGM(1,1) to electricity consumption and has examined its forecasting ability on electricity consumption using sample data from the Turkish Ministry of Energy and Natural Resources and the Asia–Pacific Economic Cooperation energy database. Experimental results demonstrate that NNGM(1,1) performs well.

Membrane-based zero liquid discharge: Myth or reality?

Abstract: The environmental impacts of brine disposal from seawater desalination plants and wastewater treatment plants represent a subject of growing concern; thus, determining the potential applicability of zero liquid discharge (ZLD) for water treatment is crucial. Membrane-based technologies are a potentially attractive strategy that can be used to reach this goal. Recent studies have highlighted that integrating a series of membrane processes is a viable approach to achieving ZLD for industrial use. However, a relatively limited number of reports have been published on the challenging problems encountered with ZLD approaches. Here, we provide a review of membrane processes that may be used in ZLD approaches and describe their problems as well as potential solutions and innovative technologies for improving their performance. Furthermore, the energy consumption of the different approaches is calculated and analyzed because it represents a major contributor to the total cost, and investments in innovative technologies are discussed. Finally, the prospects for membrane-based ZLD and further research are highlighted.

Experiencing the Analects of Confucius

Abstract: The Analects of Confucius is an important course in the curriculum of Asian Studies in the Chinese community and around the world. Students have to learn a collection of the thoughts of Confucius which have shaped world history and the soul of China. However, educators have indicated that most students fail to understand its abstract thoughts, or even realize its spirit in their daily life experience. Meanwhile, with the advancements of technology, learning with computer games is currently a rapidly developing area of interest for researchers, teachers, material writers and application developers in the educational field. Several studies have shown that by properly incorporating learning contents into game scenarios, an experiential game-based learning approach might foster students' motivation to learn through experience. In addition, the experiential game-based approach is a learning method with great potential for motivating students and stimulating their willingness to engage in continuous and constant learning. Thus, in this study, an experiential digital game has been developed and presented to a fifth grade class learning the Analects of Confucius at an elementary school in Taipei city. An experiment was conducted to evaluate the effectiveness of the proposed approach by situating the experimental group in an experiential game-based learning scenario, while the control group learned with a conventional technology-enhanced learning system. The experimental results showed that the proposed approach effectively enhanced the students' learning effects in terms of their learning motivation, deep learning strategy and acceptance of the technology. As a consequence, it is concluded that an experiential digital game-based learning approach can help students understand the conception and meaning of learning, which is important for them to become life-long learners with positive learning attitudes. An experiential game-based learning approach is proposed.A digital game is developed to enable students to experience the Analects of Confucius.An experiment was conducted to evaluate the performance of the new approach.The approach improved the students' learning motivation and deep learning strategy.The students highly accepted the new approach.

Insight into adsorption mechanism of cationic dye onto agricultural residues-derived hydrochars: Negligible role of π-π interaction

Abstract: Hydrochars derived from golden shower pod (GSH), coconut shell (CCH), and orange peel (OPH) were synthesized and applied to remove methylene green (MG5). The results indicated that the hydrochars possessed low specific surface areas (6.65-14.7m2/g), but abundant oxygen functionalities (1.69-2.12mmol/g). The hydrochars exhibited cellular and spherical morphologies. Adsorption was strongly dependent on the solution pH (2-10) and ionic strength (0-0.5M NaCl). Equilibrium can be quickly established in the kinetic study (60-120 min). The maximum Langmuir adsorption capacities at 30 °C followed the order GSH (59.6mg/g)>CCH (32.7mg/g)>OPH (15.6mg/g)> commercial glucose-prepared hydrochar (12.6mg/g). The dye adsorption efficiency was determined by the concentrations of oxygen-containing functionalities on the hydrochar surface. The adsorption process occurred spontaneously (− ΔGo) and exothermically (−ΔHo). Desorption studies confirmed the reversible adsorption process. Oxygenation of the hydrochar surface through a hydrothermal process with acrylic acid contributed to increasing MG5 adsorption and identifying the negligible role of π-π interaction to the adsorption process. The analysis of Fourier transform infrared spectrometry demonstrated that the aromatic C=C peak did not significantly decrease in intensity or shift toward higher/lower wavenumbers after adsorption, which further confirms the insignificant contribution of π-π interaction. Electrostatic attraction played a major role in adsorption mechanisms, while minor contributions were accounted for hydrogen bonding and n-π interactions. The primary adsorption mechanisms of MG5 onto hydrochar were similar to biosorbent, but dissimilar to biochar and activated carbon (i.e., π-π interaction and pore filling).

Resveratrol attenuates ICAM-1 expression and monocyte adhesiveness to TNF-α-treated endothelial cells: evidence for an anti-inflammatory cascade mediated by the miR-221/222/AMPK/p38/NF-κB pathway.

Abstract: Resveratrol, an edible polyphenolic phytoalexin, improves endothelial dysfunction and attenuates inflammation. However, the mechanisms have not been thoroughly elucidated. Therefore, we investigated the molecular basis of the effects of resveratrol on TNF-α-induced ICAM-1 expression in HUVECs. The resveratrol treatment significantly attenuated the TNF-α-induced ICAM-1 expression. The inhibition of p38 phosphorylation mediated the reduction in ICAM-1 expression caused by resveratrol. Resveratrol also decreased TNF-α-induced IκB phosphorylation and the phosphorylation, acetylation, and translocation of NF-κB p65. Moreover, resveratrol induced the AMPK phosphorylation and the SIRT1 expression in TNF-α-treated HUVECs. Furthermore, TNF-α significantly suppressed miR-221/-222 expression, which was reversed by resveratrol. miR-221/-222 overexpression decreased p38/NF-κB and ICAM-1 expression, which resulted in reduced monocyte adhesion to TNF-α-treated ECs. In a mouse model of acute TNF-α-induced inflammation, resveratrol effectively attenuated ICAM-1 expression in the aortic ECs of TNF-α-treated wild-type mice. These beneficial effects of resveratrol were lost in miR-221/222 knockout mice. Our data showed that resveratrol counteracted the TNF-α-mediated reduction in miR-221/222 expression and decreased the TNF-α-induced activation of p38 MAPK and NF-κB, thereby suppressing ICAM-1 expression and monocyte adhesion. Collectively, our results show that resveratrol attenuates endothelial inflammation by reducing ICAM-1 expression and that the protective effect was mediated partly through the miR-221/222/AMPK/p38/NF-κB pathway.

Comparisons of Self-Determination Among Students With Autism, Intellectual Disability, and Learning Disabilities: A Multivariate Analysis:

Abstract: This study examined differences in self-determination among students with autism spectrum disorders (ASD), students with intellectual disability (ID), and students with learning disabilities (LD). A total of 222 participants with an equal size group for each of the three disability categories were selected to participate in the comparison of total self-determination and domain scores. A multivariate analysis of covariance (MANCOVA) was performed on four dependent variables (DVs)/factors, including autonomy, self-regulation, psychological empowerment, and self-realization. The results indicated that students with ASD had significantly lower levels of autonomy compared with students in either other group; that students with ID had significantly lower levels of self-regulation than students with LD, but not significantly different from students with ASD; that students with ASD and students with ID had significantly lower levels of psychological empowerment than students with LD; and that students with ID had...