• Journal of the Korea Society of Computer and Information
• /
• v.18 no.7
• /
• pp.23-35
• /
• 2013

This paper proposes an effective technique that is used to automatically extract feature vectors from vibration signals for fault classification systems. Conventional mel-frequency cepstral coefficients (MFCCs) are sensitive to noise of vibration signals, degrading classification accuracy. To solve this problem, this paper proposes spectral envelope cepstral coefficients (SECC) analysis, where a 4-step filter bank based on spectral envelopes of vibration signals is used: (1) a linear predictive coding (LPC) algorithm is used to specify spectral envelopes of all faulty vibration signals, (2) all envelopes are averaged to get general spectral shape, (3) a gradient descent method is used to find extremes of the average envelope and its frequencies, (4) a non-overlapped filter is used to have centers calculated from distances between valley frequencies of the envelope. This 4-step filter bank is then used in cepstral coefficients computation to extract feature vectors. Finally, a multi-layer support vector machine (MLSVM) with various sigma values uses these special parameters to identify faulty types of induction motors. Experimental results indicate that the proposed extraction method outperforms other feature extraction algorithms, yielding more than about 99.65% of classification accuracy.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1440-1447
• /
• 2019

Currently, the prevalence of autism spectrum disorders in children is reported to be higher and shows various types of disorders. In particular, they are having difficulty in communication due to communication impairment in the area of social communication and need to be improved through training. Thus, this study proposes a method of acquiring voice information through a microphone mounted on a robot designed through preliminary research and using this information to make intelligent motions. An ANN(Artificial Neural Network) was used to classify the speech data into robot motions, and we tried to improve the accuracy by combining the Recurrent Neural Network based on Convolutional Neural Network. The preprocessing of input speech data was analyzed using MFCC(Mel-Frequency Cepstral Coefficient), and the motion of the robot was estimated using various data normalization and neural network optimization techniques. In addition, the designed ANN showed a high accuracy by conducting an experiment comparing the accuracy with the existing architecture and the method of human intervention. In order to design robot motions with higher accuracy in the future and to apply them in the treatment and education environment of children with autism.

• Particle and aerosol research
• /
• v.8 no.1
• /
• pp.17-28
• /
• 2012

Air pollution levels of gases and aerosol particles inside the Jeongneung and Hongjimun tunnels of the Naebu express way in Seoul were investigated through on-road measurement using a mobile emission laboratory (MEL) on February 8, 2011. The concentrations of $NO_x$, $CO_2$, number concentration of particles ranging 21-560 nm, and surface area of particles deposited on a human lung almost linearly increased with increasing distance from the tunnel entrance, and decreased rapidly before the tunnel exit. This trend was observed regardless of tunnel length and driving directions, which thought to be caused by semi-transverse ventilation facilities of the tunnels. The concentration increments per 1-m distance for $NO_x$, $CO_2$, deposited particle surface area, and number of particles ranging 21-560 nm were 0.61~0.80 ppb, 0.16~0.21 ppm, $0.20{\sim}0.29{\mu}m^2/cm^3$, and 117~192 particles/$cm^3$, respectively. Average pollution levels inside the two tunnels for $CO_2$, deposited particle surface area, and number of particles >5.6 nm ranged 681~748 ppm, $246{\sim}381{\mu}m^2/cm^3$, and $2.4{\sim}6.7{\times}10^5$ particles/$cm^3$, respectively. In case of $NO_x$, the maximum concentration exceeded 1 ppm. These pollution levels inside the tunnels are much higher than those at urban background sites. This result can be utilized as basic data to evaluate the effectiveness of present ventilation system for reducing the pollution level caused by vehicles inside the tunnels.

• The Journal of the Acoustical Society of Korea
• /
• v.24 no.1
• /
• pp.29-37
• /
• 2005

This paper proposes an automatic genre classification system for Korean traditional music. The Proposed system accepts and classifies queried input music as one of the six musical genres such as Royal Shrine Music, Classcal Chamber Music, Folk Song, Folk Music, Buddhist Music, Shamanist Music based on music contents. In general, content-based music genre classification consists of two stages - music feature vector extraction and Pattern classification. For feature extraction. the system extracts 58 dimensional feature vectors including spectral centroid, spectral rolloff and spectral flux based on STFT and also the coefficient domain features such as LPC, MFCC, and then these features are further optimized using SFS method. For Pattern or genre classification, k-NN, Gaussian, GMM and SVM algorithms are considered. In addition, the proposed system adopts MFC method to settle down the uncertainty problem of the system performance due to the different query Patterns (or portions). From the experimental results. we verify the successful genre classification performance over $97{\%}$ for both the k-NN and SVM classifier, however SVM classifier provides almost three times faster classification performance than the k-NN.

• Journal of Life Science
• /
• v.18 no.6
• /
• pp.884-890
• /
• 2008

Genistein, an isoflavone in soybean products, is a potential chemopreventive agent against various types of cancer. There are several studies documenting molecular alterations leading to cell cycle arrest at G2/M phase and induction of apoptosis; however, its mechanism of action and its molecular targets on the prostaglandin $E_2$ ($PGE_2$) production and telomere length regulation in human cancer remain unclear. In this study, we investigated the effect of genistein on the levels of cyclooxygenases (COXs) and telomere regulatory components of several human cancer cell lines (T24, human bladder carcinoma cells; U937, human leukemic cells; AGS, human stomach adenocarcinoma cells and SK-MEL-2, human skin melanoma cells). Genistein treatment resulted in the inhibition of cancer cell proliferation in a concentration-dependent manner. It was found that genistein treatment markedly decreased the levels of COX-2 mRNA and protein expression without significant changes in the expression of COX-1, which was correlated with a decrease in $PGE_2$ synthesis. Genistein treatment also partly inhibited the levels of human telomerase reverse transcriptase (hTERT) as well as human telomerase RNA (hTR) and telomerase-associated protein (TEP)-1, and the activity of telomerase. Taken together, these findings provide important new insights into the possible molecular mechanisms of the anti-cancer activity of genistein.

• Proceedings of the Korean Society of Applied Pharmacology
• /
• 2002.07a
• /
• pp.113-113
• /
• 2002

Phylogenetically conserved Bcl-2 family proteins play a pivotal role in the regulation of apoptosis from virus to human. Members of the Bcl-2 family consist of antiapoptotic proteins such as Bcl-2, Bcl-xL, and Bcl-w, and proapoptotic proteins such as BAD, Bax, BOD, and Bok. It has been proposed that anti- and proapoptotic Bcl-2 proteins regulate cell death by binding to each other and forming heterodimers. A delicate balance between anti- and proapoptotic Bcl-2 family members exists in each cell and the relative concentration of these two groups of proteins determines whether the cell survives or undergoes apoptosis. Mcl-1 (Myeloid cell :leukemia-1) is a member of the Bcl-2 family proteins and was originally cloned as a differentiation-induced early gene that was activated in the human myeloblastic leukemia cell line, ML-1 . Mcl-1 is expressed in a wide variety of tissues and cells including neoplastic ones. We recently identified a short splicing variant of Mcl-1 short (Mcl-IS) and designated the known Mcl-1 as Mcl-1 long (Mcl-lL). Mcl-lL protein exhibits antiapoptotic activity and possesses the BH (Bcl-2 homology) 1, BH2, BH3, and transmembrane (TM) domains found in related Bcl-2 proteins. In contrast, Mcl-1 S is a BH3 domain-only proapoptotic protein that heterodimerizes with Mcl-lL. Although both Mc1-lL and Mcl-lS proteins contain BH domains fecund in other Bcl-2 family proteins, they are distinguished by their unusually long N-terminal sequences containing PEST (proline, glutamic acid, serine, and threonine) motifs, four pairs of arginine residues, and alanine- and glycine-rich regions. In addition, the expression pattern of Mcl-1 protein is different from that of Bcl-2 suggesting a unique role (or Mcl-1 in apoptosis regulation. Tankyrasel (TRF1-interacting, ankyrin-related ADP-related polymerasel) was originally isolated based on its binding to TRF 1 (telomeric repeat binding factor-1) and contains the sterile alpha motif (SAM) module, 24 ankyrin (ANK) repeats, and the catalytic domain of poly(adenosine diphosphate-ribose) polymerase (PARP). Previous studies showed that tankyrasel promotes telomere elongation in human cells presumably by inhibiting TRFI though its poly(ADP-ribosyl)action by tankyrasel . In addition, tankyrasel poly(ADP-ribosyl)ates Insulin-responsive amino peptidase (IRAP), a resident protein of GLUT4 vesicles, and insulin stimulates the PARP activity of tankyrase1 through its phosphorylation by mitogen-activated protein kinase (MAPK). ADP-ribosylation is a posttranslational modification that usually results in a loss of protein activity presumably by enhancing protein turnover. However, little information is available regarding the physiological function(s) of tankyrase1 other than as a PARP enzyme. In the present study, we found tankyrasel as a specific-binding protein of Mcl-1 Overexpression of tankyrasel led to the inhibition of both the apoptotic activity of Mel-lS and the survival action of Mcl-lL in mammalian cells. Unlike other known tankyrasel-interacting proteins, tankyrasel did not poly(ADP-ribosyl)ate either of the Mcl-1 proteins despite its ability to decrease Mcl-1 proteins expression following coexpression. Therefore, this study provides a novel mechanism to regulate Mcl-1-modulated apoptosis in which tankyrasel downregulates the expression of Mcl-1 proteins without the involvement of its ADP-ribosylation activity.

• The Korean Journal of Physiology and Pharmacology
• /
• v.7 no.5
• /
• pp.267-273
• /
• 2003

We have investigated the effect on inducing substate(s) of positively charged residues replaced in position 172 of the second transmembrane domain in murine inward rectifier potassium channels, formed by stable or transient transfection of Kir2.1 gene in MEL or CHO cells. Single channel recordings were obtained from either cell-attached patches or inside-out patches excised into solution containing 10 mM EDTA to rule out the effect of $Mg^{2+}$ on the channel gating. The substate(s) could be recorded with all mutants D172H, D172K and D172R. The unitary current-voltage (I-V) relation was not linear with D172H at $pH_i$ 6.3, whereas the unitary I-V relation was linear at $pH_i$ 8.0. The relative occupancy at $S_{LC}$ was increased from 0.018 at $pH_i$ 8.0 to 0.45 at $pH_i$ 5.5. In H-N dimer, that was increased from 0.016 at $pH_i$ 8.0 to 0.23 at $pH_i$ 5.5. The larger the size of the side chain or $pK_a$ with mutants (D172H, D172K and D172R), the more frequent the transitions between the fully open state and substate within an opening. The conductance of the substate also depended upon the pKa or the size of the side chain. The relative occupancy at substate $S_{LC}$ with monomer D172K (0.50) was less than that in K-H dimer (0.83). However, the relative occupancy at substate with D172R (0.79) was similar to that with R-N dimer (0.82). In the contrary to ROMK1, positive charge as well as negative charge in position 172 can induce the substate rather than block the pore in murine Kir2.1. The single channel properties of the mutant, that is, unitary I-V relation, the voltage dependence of the mean open time and relative occupancy of the substates and the increased latency to the first opening, explain the intrinsic gating observed in whole cell recordings.

• Journal of the Korean Magnetics Society
• /
• v.14 no.4
• /
• pp.143-148
• /
• 2004

Effects of residual flux density M$_{res}$ and number of inspection on the detection voltage and flux density B of the gas pipeline were investigated in MFL inspection, which is widely used for the non-destructive inspection in a gas pipeline. A simulation equipment composed of the magnetizer and iron ring attached on an aluminum disc was constructed instead of a huge gas pipeline facility. With this system. the iron ring could be perfectly demagnetized and signals from the bolt screw stuck on the disc could be clearly detected so that the effects of M$_{res}$S and the inspection number on the detection voltage and B of iron ring were effectively investigated. With increasing the number of inspection, M$_{res}$, B of the iron ring and the detection voltage decreased and then kept at constant values while final M$_{res}$ increased with increasing initial M$_{res}$. If inspection condition were kept unchanged, the detection voltage was proportional to the last M$_{res}$ of the iron ring instead of B. This was probably due to magnetic hysteresis of the iron ring inherited from magnetic domain so that consideration on the magnetic hysteresis was inevitable in the analysis of MFL signal from defects of a gas pipeline. A new inspection scheme using the magnetizer with reversed magnetization in the subsequent inspection was proposed from the result that a high detection voltage could be obtained in the first inspection of gas pipeline with positive M$_{res}$.

• The Korean Journal of Financial Management
• /
• v.7 no.1
• /
• pp.1-20
• /
• 1990

In most of studies on market efficiency, the stability of risk measures and the normality of residuals unexplained by the pricing model are presumed. This paper re-examines stock splits, taking the possible violation of two assumptions into accounts. The results does not change the previous studies. But, the size of excess returns during the 2-week period before announcements decreases by 43%. The results also support that betas change around announcements and the serial autocorrelation of residuals is caused by events. Based on the results, the existing excess returns are most likely explained as a compensation to old shareholders for unwanted risk increases in their portfolio, or by uses of incorrect betas in testing models. In addition, the model suggested in the paper provides a measure for the speed of adjustment of the market to the new information arrival and the intensity of information contents.

• The Korean Journal of Applied Statistics
• /
• v.32 no.2
• /
• pp.213-227
• /
• 2019

In this paper, we propose an end-to-end deep learning model combining Bayesian neural network with Korean speech recognition. In the past, Korean speech recognition was a complicated task due to the excessive parameters of many intermediate steps and needs for Korean expertise knowledge. Fortunately, Korean speech recognition becomes manageable with the aid of recent breakthroughs in "End-to-end" model. The end-to-end model decodes mel-frequency cepstral coefficients directly as text without any intermediate processes. Especially, Connectionist Temporal Classification loss and Attention based model are a kind of the end-to-end. In addition, we combine Bayesian neural network to implement the end-to-end model and obtain Monte Carlo estimates. Finally, we carry out our experiments on the "WorimalSam" online dictionary dataset. We obtain 4.58% Word Error Rate showing improved results compared to Google and Naver API.