• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.16 no.4
• /
• pp.153-163
• /
• 2017

This research is based on the previous research that personally preferred safe distance, rotating angle and speed are differentiated. Thus, we use machine learning model for recognizing maneuvering modes trained per personal or per similar driving pattern groups, and we evaluate automatic driving according to maneuvering modes. By utilizing driving knowledge, we subdivided 8 kinds of longitudinal modes and 4 kinds of lateral modes, and by combining the longitudinal and lateral modes, we build 21 kinds of maneuvering modes. we train the labeled data set per time stamp through RNN, LSTM and Bi-LSTM models by the trips of drivers, which are supervised deep learning models, and evaluate the maneuvering modes of automatic driving for the test data set. The evaluation dataset is aggregated of living trips of 3,000 populations by VTTI in USA for 3 years and we use 1500 trips of 22 people and training, validation and test dataset ratio is 80%, 10% and 10%, respectively. For recognizing longitudinal 8 kinds of maneuvering modes, RNN achieves better accuracy compared to LSTM, Bi-LSTM. However, Bi-LSTM improves the accuracy in recognizing 21 kinds of longitudinal and lateral maneuvering modes in comparison with RNN and LSTM as 1.54% and 0.47%, respectively.

• Journal of the Korea Society of Computer and Information
• /
• v.15 no.8
• /
• pp.13-21
• /
• 2010

One of the most distinguished features of the DHP association rules mining algorithm is that it counts the support of hash key combinations composed of k items at phase k-1, and uses the counted support for pruning candidate large itemsets to improve performance. At this time, it is desirable for each hash key combination to have a separate count variable, where it is impossible to allocate the variables owing to memory shortage. So, the algorithm uses a direct hashing mechanism in which several hash key combinations conflict and are counted in a same hash bucket. But the direct hashing mechanism is not efficient because the distribution of hash key combinations is unvalanced by the characteristics sourced from the mining process. This paper proposes a mapped perfect hashing function which maps the region of hash key combinations into a continuous integer space for phase 3 and maximizes the efficiency of direct hashing mechanism. The results of a performance test experimented on 42 test data sets shows that the average performance improvement of the proposed hashing mechanism is 7.3% compared to the existing method, and the highest performance improvement is 16.9%. Also, it shows that the proposed method is more efficient in case the length of transactions or large itemsets are long or the number of total items is large.

• Publications of The Korean Astronomical Society
• /
• v.21 no.2
• /
• pp.67-74
• /
• 2006

We have developed a control electronics system for an infrared detector array of KASINICS (KASI Near Infrared Camera System), which is a new ground-based instrument of the Korea Astronomy and Space science Institute (KASI). Equipped with a $512{\times}512$ InSb array (ALADDIN III Quadrant, manufactured by Raytheon) sensitive from 1 to $5{\mu}m$, KASINICS will be used at J, H, Ks, and L-bands. The controller consists of DSP(Digital Signal Processor), Bias, Clock, and Video boards which are installed on a single VME-bus backplane. TMS320C6713DSP, FPGA(Field Programmable Gate Array), and 384-MB SDRAM(Synchronous Dynamic Random Access Memory) are included in the DSP board. DSP board manages entire electronics system, generates digital clock patterns and communicates with a PC using USB 2.0 interface. The clock patterns are downloaded from a PC and stored on the FPGA. UART is used for the communication with peripherals. Video board has 4 channel ADC which converts video signal into 16-bit digital numbers. Two video boards are installed on the controller for ALADDIN array. The Bias board provides 16 dc bias voltages and the Clock board has 15 clock channels. We have also coded a DSP firmware and a test version of control software in C-language. The controller is flexible enough to operate a wide range of IR array and CCD. Operational tests of the controller have been successfully finished using a test ROIC (Read-Out Integrated Circuit).

• Journal of Life Science
• /
• v.19 no.7
• /
• pp.917-922
• /
• 2009

The effect of adding docosahexaenoic acid into an n-3 fatty acid adequate diet on the improvement of learning related brain function was investigated. On the second day after conception, Sprague Dawley strain dams were subjected to a diet containing either n-3 fatty acid adequate (Adq, 3.4% linolenic acid) or n-3 fatty acid adequate+docosahexaenoic acid (Adq+DHA, 3.31%linolenic acid plus 9.65% DHA). After weaning, male pups were fed on the same diet of their respective dams until adulthood. Motor activity and Morris water maze tests were measured at 10 weeks. In the motor activity test, there were no statistically significant differences in moving time and moving distance between the Adq and Adq+DHA diet groups. The n-3 fatty acid adequate with DHA (Adq+DHA) group tended to show a shorter escape latency, swimming time and swimming distance compared to the n-3 fatty acid adequate group (Adq), but the differences were not statistically significant. There was no difference in resting time, but the Adq+DHA group showed a higher swimming speed compared to the Adq group. In memory retention trials, the numbers of crossing of the platform position (region A), in which the hidden platform was placed, were significantly greater than those of other regions for both Adq and Adq+DHA groups. Based on these results, adding DHA into the n-3 fatty acid adequate diet from gestation to adulthood tended to induce better spatial learning performance in Sprague Dawley rats as assessed by the Morris water maze test, although the difference was not significant.

• Nutrition Research and Practice
• /
• v.12 no.3
• /
• pp.199-207
• /
• 2018

BACKGROUND/OBJECTIVES: Fermented Laminaria japonica (FL), a type sea tangle used as a functional food ingredient, has been reported to possess cognitive improving properties that may aid in the treatment of common neurodegenerative disorders, such as dementia. MATERIALS/METHODS: We examined the effects of FL on scopolamine (Sco)- and ethanol (EtOH)-induced hippocampus-dependent memory impairment, using the Passive avoidance (PA) and Morris water maze (MWM) tests. To examine the underlying mechanisms associated with neuroprotective effects, we analyzed acetylcholine (ACh) and acetylcholinesterase (AChE) activity, brain tissue expression of muscarinic acetylcholine receptor (mAChR), cAMP response element binding protein (CREB) and extracellular signal-regulated kinases 1/2 (ERK1/2), and immunohistochemical analysis, in the hippocampus of mice, compared to current drug therapy intervention. Biochemical blood analysis was carried out to determine the effects of FL on alanine transaminase (ALT), aspartate transaminase (AST), and triglyceride (TG) and total cholesterol (TC) levels. 7 groups (n = 10) consisted of a control (CON), 3 Sco-induced dementia and 3 EtOH-induced dementia groups, with both dementia group types containing an untreated group (Sco and EtOH); a positive control, orally administered donepezil (Dpz) (4mg/kg) (Sco + Dpz and EtOH + Dpz); and an FL (50 mg/kg) treatment group (Sco + FL50 and EtOH + FL50), orally administered over the 4-week experimental period. RESULTS: FL50 significantly reduced EtOH-induced increase in AST and ALT levels. FL50 treatment reduced EtOH-impaired step-through latency time in the PA test, and Sco- and EtOH-induced dementia escape latency times in the MWM test. Moreover, anticholinergic effects of Sco and EtOH on the brain were reversed by FL50, through the attenuation of AChE activity and elevation of ACh concentration. FL50 elevated ERK1/2 protein expression and increased p-CREB (ser133) in hippocampus brain tissue, according to Western blot and immunohistochemistry analysis, respectively. CONCLUSION: Overall, these results suggest that FL may be considered an efficacious intervention for Sco- and EtOH-induced dementia, in terms of reversing cognitive impairment and neuroplastic dysfunction.

• Journal of Applied Biological Chemistry
• /
• v.64 no.3
• /
• pp.299-307
• /
• 2021

Accumulation of amyloid beta (Aβ) and oxidative stress are the most common reason of Alzheimer's disease (AD). In the present study, we investigated the cognitive improvement effects of butanol (BuOH) fraction from Momordica charantia in Aβ_25-35-induced AD mouse model. To develop an AD mouse model, mice were received injection of Aβ_25-35, and then orally administered BuOH fraction from M. charantia at doses of 100 and 200 mg/kg/day during 14 days. In the T-maze and novel object recognition test, administration of BuOH fraction from M. charantia L. at doses of 100 and 200 mg/kg/day improved spatial ability and novel object recognition by increased explorations of novel route and new object. In addition, BuOH fraction of M. charantia-administered groups improved learning and memory abilities by decreased time to reach hidden platform in Morris water maze test. Oral administration of BuOH fraction from M. charantia significantly inhibited lipid peroxidation and nitric oxide levels in the brain, liver, and kidney compared with Aβ_25-35-induced control group. These results indicated that BuOH fraction of M. charantia improved Aβ_25-35-induced cognitive impairment by attenuating oxidative stress. Therefore, M. charantia could be useful for protection from Aβ_25-35-induced cognitive impairment.

• Journal of Advanced Navigation Technology
• /
• v.23 no.1
• /
• pp.84-89
• /
• 2019

This is a pilot study of machine learning based structural health monitoring system using flight data of composite aircraft. In this study, the most suitable machine learning algorithm for structural health monitoring was selected and dimensionality reduction method for application on the actual flight data was conducted. For these tasks, impact test on the cantilever beam with added mass, which is the simulation of damage in the aircraft wing structure was conducted and classification model for damage states (damage location and level) was trained. Through vibration test of cantilever beam with fiber bragg grating (FBG) sensor, data of normal and 12 damaged states were acquired, and the most suitable algorithm was selected through comparison between algorithms like tree, discriminant, support vector machine (SVM), kNN, ensemble. Besides, through neighborhood component analysis (NCA) feature selection, dimensionality reduction which is necessary to deal with high dimensional flight data was conducted. As a result, quadratic SVMs performed best with 98.7% for without NCA and 95.9% for with NCA. It is also shown that the application of NCA improved prediction speed, training time, and model memory.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.3
• /
• pp.287-301
• /
• 2022

The present study aims to investigate the difference of soundscape evaluation results from on-site field test and laboratory test which are commonly used for soundscape surveys. In order to do this, both field and lab tests were carried out at four different areas in Cheongju city. On-site questionnaire surveys were undertaken to 65 people at 13 points. Laboratory listening tests were carried out to 48 adults using recorded sounds and video. Laboratory tests were undertaken to two different groups who had experience of field survey or not. Also, two different sound reproduction tools, headphones and speakers, were used in laboratory tests. As a result, it was found that there is a very close correlation between sound loudness and annoyance in both field and laboratory tests. However, it was concluded that there must be a difference in recognizing the figure sounds between field and laboratory tests since it is hard to apprehend on-site situation only using visual and aural information provided in laboratory tests. In laboratory tests, it was shown that there is a some difference in perceived most loud figure sounds in two groups using headphones and speakers. Also, it was analyzed that there is a tendency that field experienced people recognize the figure sounds using their experienced memory while non-experienced people can not perceive the figure sounds.

• Korean Journal of Psychosomatic Medicine
• /
• v.27 no.2
• /
• pp.155-163
• /
• 2019

Objectives : The aim of this study was to compare severity, neurocognitive functions, and behavioral and psychological symptoms of dementia (BPSD) according to the degree of temporal lobe atrophy (MTA) in Korean patients with dementia due to Alzheimer's disease and mild cognitive impairment due to Alzheimer's disease. Methods : Participants were 114 elderly subjects diagnosed with Alzheimer's disease or mild cognitive impairment in this cross-sectional study. MTA in brain MRI was rated with standardized visual rating scales (Scheltens scale) and the subjects were divided into two groups according to Scheltens scale. Severity was evaluated with Clinical Dementia Rating (CDR) and Global Deterioration Scale (GDS). Neurocognitive functions was evaluated with the Korean version of Short Blessed Test (SBT-K) and the Korean version of the Consortium to Establish a Registry for Alzheimer's Disease assessment packet (CERAD-K). BPSD was evaluated with the Korean version of the Neuropsychiatric Inventory (K-NPI). Independent t-test was performed to compare severity, neurocognitive functions, and BPSD between two groups. Results : The group with high severity of MTA showed significantly lower scores in CDR, SBT-K, MMSE-KC, modified Boston naming test, word list recognition, and word list memory (p<0.05). There were no differences in K-NPI scores between two groups. Conclusions : Severity and neurocognitive functions of dementia had significant positive association with MTA, but BPSD had no association with MTA. Evaluating MTA seems to have potential benefit in diagnosing and treating neurocognitive impairments in the elderly. Further evaluation is needed to confirm the association between certain brain structures and BPSD.

• Journal of Intelligence and Information Systems
• /
• v.21 no.1
• /
• pp.119-142
• /
• 2015

With the explosive growth in the volume of information, Internet users are experiencing considerable difficulties in obtaining necessary information online. Against this backdrop, ever-greater importance is being placed on a recommender system that provides information catered to user preferences and tastes in an attempt to address issues associated with information overload. To this end, a number of techniques have been proposed, including content-based filtering (CBF), demographic filtering (DF) and collaborative filtering (CF). Among them, CBF and DF require external information and thus cannot be applied to a variety of domains. CF, on the other hand, is widely used since it is relatively free from the domain constraint. The CF technique is broadly classified into memory-based CF, model-based CF and hybrid CF. Model-based CF addresses the drawbacks of CF by considering the Bayesian model, clustering model or dependency network model. This filtering technique not only improves the sparsity and scalability issues but also boosts predictive performance. However, it involves expensive model-building and results in a tradeoff between performance and scalability. Such tradeoff is attributed to reduced coverage, which is a type of sparsity issues. In addition, expensive model-building may lead to performance instability since changes in the domain environment cannot be immediately incorporated into the model due to high costs involved. Cumulative changes in the domain environment that have failed to be reflected eventually undermine system performance. This study incorporates the Markov model of transition probabilities and the concept of fuzzy clustering with CBCF to propose predictive clustering-based CF (PCCF) that solves the issues of reduced coverage and of unstable performance. The method improves performance instability by tracking the changes in user preferences and bridging the gap between the static model and dynamic users. Furthermore, the issue of reduced coverage also improves by expanding the coverage based on transition probabilities and clustering probabilities. The proposed method consists of four processes. First, user preferences are normalized in preference clustering. Second, changes in user preferences are detected from review score entries during preference transition detection. Third, user propensities are normalized using patterns of changes (propensities) in user preferences in propensity clustering. Lastly, the preference prediction model is developed to predict user preferences for items during preference prediction. The proposed method has been validated by testing the robustness of performance instability and scalability-performance tradeoff. The initial test compared and analyzed the performance of individual recommender systems each enabled by IBCF, CBCF, ICFEC and PCCF under an environment where data sparsity had been minimized. The following test adjusted the optimal number of clusters in CBCF, ICFEC and PCCF for a comparative analysis of subsequent changes in the system performance. The test results revealed that the suggested method produced insignificant improvement in performance in comparison with the existing techniques. In addition, it failed to achieve significant improvement in the standard deviation that indicates the degree of data fluctuation. Notwithstanding, it resulted in marked improvement over the existing techniques in terms of range that indicates the level of performance fluctuation. The level of performance fluctuation before and after the model generation improved by 51.31% in the initial test. Then in the following test, there has been 36.05% improvement in the level of performance fluctuation driven by the changes in the number of clusters. This signifies that the proposed method, despite the slight performance improvement, clearly offers better performance stability compared to the existing techniques. Further research on this study will be directed toward enhancing the recommendation performance that failed to demonstrate significant improvement over the existing techniques. The future research will consider the introduction of a high-dimensional parameter-free clustering algorithm or deep learning-based model in order to improve performance in recommendations.