• Journal of Intelligence and Information Systems
• /
• v.19 no.4
• /
• pp.123-132
• /
• 2013

As the smartphones are equipped with various sensors such as the accelerometer, GPS, gravity sensor, gyros, ambient light sensor, proximity sensor, and so on, there have been many research works on making use of these sensors to create valuable applications. Human activity recognition is one such application that is motivated by various welfare applications such as the support for the elderly, measurement of calorie consumption, analysis of lifestyles, analysis of exercise patterns, and so on. One of the challenges faced when using the smartphone sensors for activity recognition is that the number of sensors used should be minimized to save the battery power. When the number of sensors used are restricted, it is difficult to realize a highly accurate activity recognizer or a classifier because it is hard to distinguish between subtly different activities relying on only limited information. The difficulty gets especially severe when the number of different activity classes to be distinguished is very large. In this paper, we show that a fairly accurate classifier can be built that can distinguish ten different activities by using only a single sensor data, i.e., the smartphone accelerometer data. The approach that we take to dealing with this ten-class problem is to use the ensemble of nested dichotomy (END) method that transforms a multi-class problem into multiple two-class problems. END builds a committee of binary classifiers in a nested fashion using a binary tree. At the root of the binary tree, the set of all the classes are split into two subsets of classes by using a binary classifier. At a child node of the tree, a subset of classes is again split into two smaller subsets by using another binary classifier. Continuing in this way, we can obtain a binary tree where each leaf node contains a single class. This binary tree can be viewed as a nested dichotomy that can make multi-class predictions. Depending on how a set of classes are split into two subsets at each node, the final tree that we obtain can be different. Since there can be some classes that are correlated, a particular tree may perform better than the others. However, we can hardly identify the best tree without deep domain knowledge. The END method copes with this problem by building multiple dichotomy trees randomly during learning, and then combining the predictions made by each tree during classification. The END method is generally known to perform well even when the base learner is unable to model complex decision boundaries As the base classifier at each node of the dichotomy, we have used another ensemble classifier called the random forest. A random forest is built by repeatedly generating a decision tree each time with a different random subset of features using a bootstrap sample. By combining bagging with random feature subset selection, a random forest enjoys the advantage of having more diverse ensemble members than a simple bagging. As an overall result, our ensemble of nested dichotomy can actually be seen as a committee of committees of decision trees that can deal with a multi-class problem with high accuracy. The ten classes of activities that we distinguish in this paper are 'Sitting', 'Standing', 'Walking', 'Running', 'Walking Uphill', 'Walking Downhill', 'Running Uphill', 'Running Downhill', 'Falling', and 'Hobbling'. The features used for classifying these activities include not only the magnitude of acceleration vector at each time point but also the maximum, the minimum, and the standard deviation of vector magnitude within a time window of the last 2 seconds, etc. For experiments to compare the performance of END with those of other methods, the accelerometer data has been collected at every 0.1 second for 2 minutes for each activity from 5 volunteers. Among these 5,900 ($=5{\times}(60{\times}2-2)/0.1$) data collected for each activity (the data for the first 2 seconds are trashed because they do not have time window data), 4,700 have been used for training and the rest for testing. Although 'Walking Uphill' is often confused with some other similar activities, END has been found to classify all of the ten activities with a fairly high accuracy of 98.4%. On the other hand, the accuracies achieved by a decision tree, a k-nearest neighbor, and a one-versus-rest support vector machine have been observed as 97.6%, 96.5%, and 97.6%, respectively.

• Restorative Dentistry and Endodontics
• /
• v.25 no.4
• /
• pp.527-535
• /
• 2000

The purpose of this study was to evaluate effects of smear layer and dentin primers on the sealing ability of root canals. 126 extracted human teeth with single, straight canals and mature apices were used. The Samples were first classified into six groups as follows: presence of smear layer; absence of smear layer; Scotchbond Bond Multi-Purpose; All Bond 2; Mac Bond 2; Clearfil Liner Bond 2. A Positive control was also established. All teeth except the control group were then obturated with thermoplasticized gutta-percha and AH26. Electrochemical and dye penetration technique were later used to evaluate the degree of micro leakage through the root canal. Seventy teeth were then immersed in a 1% potassium chloride solution and An external power supply(DC 10 V) was then applied to the circuit for the electrochemical microleakage test. The degree of Microleakage was determined over period of 28 days before being evaluated. In total, 48 teeth were submitted to the dye infiltration technique. All specimen were suspended in 2% methylene blue dye for 1 week before being longitudinally split. The degree of dye infiltration was measured under a stereo microscope at ${\times}10$ magnification and evaluated. The results were as follows: 1. Apical microleakage increased throughout the test period in all group and one group having a smear layer showed a dramatic increase under electrochemical test (p<0.05). In the group having smear layer, the degree of apical microleakage was the highest, and the micro leakage was much higher than in the smear layer removed group in electrochemical test (p<0.05). Scotchbond Multi-Purpose, All Bond 2, Mac Bond 2 and Clearfil Liner Bond 2 showed lower micro leakage than one group having smear layer. The All Bond 2 and Clearfil Liner Bond 2 treated groups showed the lowest microleakage in electrochemical test (p<0.05). 2. There was no significant difference between the experimental groups in dye penetration technique. These results suggested that the removal of the smear layer from root canal and concomitantly the application of dentin primer into root canal could improve the sealing ability of root canal obturation.

• Korean journal of food and cookery science
• /
• v.28 no.6
• /
• pp.805-811
• /
• 2012

This study was carried out in order to optimize the manufacturing condition of fried garlic flakes as well as to investigate the physicochemical properties of the flakes. Fried garlic flake samples were prepared as follows: garlic was sliced by a thickness of 1.5 mm, 2.0 mm, 2.5 mm, which were measured by a thickness gage. The samples were fried in vegetable oil under different temperatures of $140{\sim}150^{\circ}C$, $160{\sim}170^{\circ}C$ and $180{\sim}185^{\circ}C$. The compression strength depending on the height (h) was measured in order to find the thickness effect by the rheometer (force control: 50 N, h: 3.25 mm). Moreover, the sample with 1.5 mm thickness showed crisp phenomena of the split compared with the crush shape of the 2.0 mm and 2.5 mm thick samples. The result of strength for time dependence showed a sample with a thickness of 1.5 mm, which was measured 5~9 times more than the 2.0 mm and 2.5 mm thick samples. We thought the reason that the 1.5 mm sample had less response power equivalent to compression force than the other samples. Alliin has been found to affect the immune responses in the blood, it is a derivative of the amino acid cysteine and is also quite heat stable. The LC system with a UV detection at 210 nm consists of a separation on a Zorbax TMS column and isocratic elution with water and ACN as a mobile phase. The alliin contents of raw and fried garlic flake under $140{\sim}150^{\circ}C$, $160{\sim}170^{\circ}C$ and $180{\sim}185^{\circ}C$ were 18.10 mg/mL, 14.0 mg/mL, 11.6 mg/mL and 11.1 mg/mL, respectively. The decrement of alliin content under different temperature was a small quantity hence, we confirmed that the increasing manufacturing temperature was not affected by the alliin content. Examining for the particle structure of fried garlic flakes by a polarization microscope, the color of the sample treated at $160{\sim}170^{\circ}C$ was pure yellow. Furder, the fiber shaped particle, which has an effect on the tough texture, almost did not appear compared to the different temperature conditions. Finally, the sensory test for the preference of fried garlic flake under different conditions was carried out and the scores for various sensory characteristics were surveyed. According to the physicochemical measurements and sensory evaluation, we confirmed that the optimum manufacturing condition of fried garlic flake was 1.5 mm thick at a temperature of $160{\sim}170^{\circ}C$.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2011.08a
• /
• pp.148-148
• /
• 2011

Recently, there have been many research activities to develop the large-area plasma source, which is able to generate the high-density plasma with relatively good uniformity, for the plasma processing in the thin-film solar cell and display panel industries. The large-area CCP sources have been applied to the PECVD process as well as the etching. Especially, the PECVD processes for the depositions of various films such as a-Si:H, ${\mu}c$-Si:H, Si3N4, and SiO2 take a significant portion of processes. In order to achieve higher deposition rate (DR), good uniformity in large-area reactor, and good film quality (low defect density, high film strength, etc.), the application of VHF (>40 MHz) CCP is indispensible. However, the electromagnetic wave effect in the VHF CCP becomes an issue to resolve for the achievement of good uniformity of plasma and film. Here, we propose a new electrode as part of a method to resolve the standing wave effect in the large-area VHF CCP. The electrode is split up a series of strip-type electrodes and the strip-type electrodes and the ground ones are arranged by turns. The standing wave effect in the longitudinal direction of the strip-type electrode is reduced by using the multi-feeding method of VHF power and the uniformity in the transverse direction of the electrodes is achieved by controlling the gas flow and the gap length between the powered electrodes and the substrate. Also, we provide the process results for the growths of the a-Si:H and the ${\mu}c$-Si:H films. The high DR (2.4 nm/s for a-Si:H film and 1.5 nm/s for the ${\mu}c$-Si:H film), the controllable crystallinity (~70%) for the ${\mu}c$-Si:H film, and the relatively good uniformity (1% for a-Si:H film and 7% for the ${\mu}c$-Si:H film) can be obtained at the high frequency of 40 MHz in the large-area discharge (280 mm${\times}$540 mm). Finally, we will discuss the issues in expanding the multi-electrode to the 8G class large-area plasma processing (2.2 m${\times}$2.4 m) and in improving the process efficiency.

• Korean Journal of Optics and Photonics
• /
• v.16 no.3
• /
• pp.248-253
• /
• 2005

We designed and fabricated a compact multi-mode interference (MMI) wavelength demultiplexer using the concept of 'Improved Quasi-State' modes. The output power and extinction ratio were improved by utilizing modal phase error which is specially occurred in low-index contrast. For a designed demultiplexer, the mode propagation analysis with effective index approximation shows significant improvement of extinction ratio to -25 dB for both $1.31{\mu}m\;and\;1.51{\mu}m$ wavelength region and the split-length was reduced about 1/5 of other MMI devices. The fabricated device shows successful characteristics for both 1.31 and $1.55{\mu}m$ wavelengths. These results demonstrate the potential of low-index materials system and the embossing process for photonic integrated circuits.

• The KIPS Transactions:PartD
• /
• v.15D no.6
• /
• pp.755-766
• /
• 2008

Recently, NAND flash memory is becoming into the spotlight as a next-generation storage device because of its small size, fast speed, low power consumption, and etc. compared to the hard disk. However, due to the distinct characteristics such as erase-before-write architecture, asymmetric operation speed and unit, disk-based systems and applications may result in severe performance degradation when directly implementing them on NAND flash memory. Especially when a B-tree is implemented on NAND flash memory, intensive overwrite operations may be caused by record inserting, deleting, and reorganizing. These may result in severe performance degradation. Although ${\mu}$-tree has been proposed in order to overcome this problem, it suffers from frequent node split and rapid increment of its height. In this paper, we propose Log-Structured B-Tree(LSB-Tree) where the corresponding log node to a leaf node is allocated for update operation and then the modified data in the log node is stored at only one write operation. LSB-tree reduces additional write operations by deferring the change of parent nodes. Also, it reduces the write operation by switching a log node to a new leaf node when inserting the data sequentially by the key order. Finally, we show that LSB-tree yields a better performance on NAND flash memory by comparing it to ${\mu}$-tree through various experiments.

• The Southeast Asian review
• /
• v.27 no.1
• /
• pp.131-161
• /
• 2017

The political dynamics of Malaysia in 2016 should be seen as a process of losing an political opportunity mainly due to the split in opposition parties. The opportunity for political development was triggered by the ruling party in crisis. The ongoing 1Malaysia Development Berhad (1MDB) scandal involving the Prime Minister Najib Razak would have provided a favorable condition for the transfer of power. The opposition parties have however failed to utilize the chance that has arisen since the general elections in 2008 and 2013 due to the chronic problem of disunity. It can be seen as distortions of political development referring to a phenomenon in which a chance for regime change formed by the crisis in authoritarian regime is distorted by internal conflicts among opposition parties. Malaysia's political turmoil seemed to paralyze its economy while foreign policy was used as a tool for domestic politics. It was reported that the key economic indicator have worsen including exports and budget deficit. The ringgit had dropped to its lowest level since the economic crisis in 1997-98 which was mainly attributed to diminishing credibility on the Najib's administration. Najib's political struggle has also impeded Malaysia's foreign policy which has attempt to embrace China and the Rohingya issue. The chance to manage key risks would be diminished if oppositions' disunity continues as there is speculation that the general election could be held in 2017.

• Anxiety and mood
• /
• v.15 no.1
• /
• pp.53-60
• /
• 2019

Objective : GRIT is a non-cognitive trait which is defined as perseverance and passion for long-term goals. It predicts success, performance and thedifference from other traits. The purpose of this study was to examine the reliability and validity of the Korean version of the GRIT scale. Methods : A total of 92 patients were enrolled in the study. All patients received psychiatric assessment including Clinical Useful Depression Outcome Scale (CUDOS), Clinical Useful Anxiety Outcome Scale (CUXOS), Patient Health Questionnaire (PHQ-15), Connor-Davidson Resilience Scale (CDRS), Brief Resilience Scale (BRS), and GRIT as well as demographic assessment. Cronbach's alpha coefficient of total GRIT score and the split-half reliability of each item was calculated to assess test reliability. Exploratory and confirmatory factor analyses were performed to select the best fitting model and assess construct validity. Finally, a correlation analysis was performed to check convergent and discriminant validity. Results : Cronbach's alpha coefficient for GRIT was found to be 0.85 and all Cronbach's alpha were more than 0.8 even in cases where all items were deleted. We found 3 appropriate factor models in exploratory factor analysis, compared them with 3 models and chose the 2-factor model as the most suitable based on the best fit test. Finally, correlation of the GRIT with CUDOS, CUXOS, PHQ-15, CDRS and BRS were statistically significant (all p<0.01), with relatively low correlation coefficient. Conclusion : This study indicates that the Korean version of GRIT is a reliable and valid instrument for investigating individual power of passion and perseverance.

• Journal of the Korean Geotechnical Society
• /
• v.22 no.4
• /
• pp.41-49
• /
• 2006

This paper demonstrates the feasibility of using shallow S-wave, high-resolution seismic reflection surveys to characterize geological structure and stratigraphy of basement rocks for civil engineering purposes. S-wave seismic reflections from depths less than 20 m were recorded along the top of steep readout slopes. Seismic reflection data were recorded using a standard CDP acquisition method with a 24-channel seismograph and a sledge-hammer SH-wave source. The data were acquired using a split-spread source-receiver geometry with a 2 m shot-and-receiver interval, and then were processed to enhance S/N ratio of the data, to improve resolvable power of the seismic section, and to get velocity information of the basement rock. The final seismic reflection profiles using the CDP technique has imaged surfaces as shallow as less than 1m and resolved beds as thin as 1m. The migrated reflection sections possess sufficient quality to correlate the prominent reflection events to the bedding planes and faults identified on the readout outcrops. Similar S-wave reflection surveys could also be used to produce the necessary details of a geological structure of shallow bedrocks to pinpoint optimum locations for monitor wells of civil engineering purposes.

• Korean Journal of Heritage: History & Science
• /
• v.52 no.4
• /
• pp.170-183
• /
• 2019

Fiber scutching refers to the process of extracting fibers from plants by separating or extracting fibers from the raw materials. As the definition of the term implies, the "Fiber Scutching" is performed on plants with advanced bast fiber as the primary material processing technique performed on plant materials. Some of the most popular phosphorus plants are ramie, hemp, flax, and the paper mulberry, which have a long history of cultivation and a wide range of distribution, making them very universal as a material supporting human life and culture. This study was described in Oju-yeonmunjangjeon-sango but was designed to re-examine the method of breaking dried stalks, which is currently unused in Korea, to examine the feasibility and characteristics of the technology. As a result of sampling and experimenting with hemp bast using the method recorded in the literature, hemp fiber was actually produced. The criteria for removing the shell from the hemp stem were the degree of discoloration and drying, and only when the stalk was completely discolored to yellow could segregation of the stalk from the shell be performed. The amount of sunlight and temperature were conditions that accelerated drying. However, if exposed for a long time, it is confirmed that hemp bast will be in a suitable condition to process, regardless of the amount of sunlight and temperature. 'Breaking of dried stalks', which utilizes the physical power of 'threshing with a flail' is considered a core process of the fiber scutching technique in 'Yukjin' in Hamgyeong-do. The bark and the core of the hemp were separated by tapping, the bast was thinly split, and the shell was peeled off, making it suitable for collecting with thread. The method of collecting the fibers by applying physical power causes downing on the fibers, which is to be generally avoided in the manufacture of bast fabric woven hemp or ramie. However, Hamgyeong-do's fiber scutching method seems to have applied this principle to the method of making fragile fabrics by using it in reverse. This method is distinct from the steaming or boiling of the stalks' in Andong, Korea, and it is similar to the Western method of spinning fabrics.