• Title/Summary/Keyword: Gradient analysis

Search Result 2,350, Processing Time 0.034 seconds

Prediction Model of User Physical Activity using Data Characteristics-based Long Short-term Memory Recurrent Neural Networks

  • Kim, Joo-Chang;Chung, Kyungyong
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.13 no.4
    • /
    • pp.2060-2077
    • /
    • 2019
  • Recently, mobile healthcare services have attracted significant attention because of the emerging development and supply of diverse wearable devices. Smartwatches and health bands are the most common type of mobile-based wearable devices and their market size is increasing considerably. However, simple value comparisons based on accumulated data have revealed certain problems, such as the standardized nature of health management and the lack of personalized health management service models. The convergence of information technology (IT) and biotechnology (BT) has shifted the medical paradigm from continuous health management and disease prevention to the development of a system that can be used to provide ground-based medical services regardless of the user's location. Moreover, the IT-BT convergence has necessitated the development of lifestyle improvement models and services that utilize big data analysis and machine learning to provide mobile healthcare-based personal health management and disease prevention information. Users' health data, which are specific as they change over time, are collected by different means according to the users' lifestyle and surrounding circumstances. In this paper, we propose a prediction model of user physical activity that uses data characteristics-based long short-term memory (DC-LSTM) recurrent neural networks (RNNs). To provide personalized services, the characteristics and surrounding circumstances of data collectable from mobile host devices were considered in the selection of variables for the model. The data characteristics considered were ease of collection, which represents whether or not variables are collectable, and frequency of occurrence, which represents whether or not changes made to input values constitute significant variables in terms of activity. The variables selected for providing personalized services were activity, weather, temperature, mean daily temperature, humidity, UV, fine dust, asthma and lung disease probability index, skin disease probability index, cadence, travel distance, mean heart rate, and sleep hours. The selected variables were classified according to the data characteristics. To predict activity, an LSTM RNN was built that uses the classified variables as input data and learns the dynamic characteristics of time series data. LSTM RNNs resolve the vanishing gradient problem that occurs in existing RNNs. They are classified into three different types according to data characteristics and constructed through connections among the LSTMs. The constructed neural network learns training data and predicts user activity. To evaluate the proposed model, the root mean square error (RMSE) was used in the performance evaluation of the user physical activity prediction method for which an autoregressive integrated moving average (ARIMA) model, a convolutional neural network (CNN), and an RNN were used. The results show that the proposed DC-LSTM RNN method yields an excellent mean RMSE value of 0.616. The proposed method is used for predicting significant activity considering the surrounding circumstances and user status utilizing the existing standardized activity prediction services. It can also be used to predict user physical activity and provide personalized healthcare based on the data collectable from mobile host devices.

A Study on Measuring Urban Sprawl and Its Policy Implications for Urban Growth Management and Urban Regeneration in Seoul Capital Region (수도권 도시 스프롤 평가에 따른 도시성장관리 및 도시재생 정책 방향에 관한 연구)

  • Jeon, Hye-Jin;Woo, Myungje
    • Journal of the Korean Regional Science Association
    • /
    • v.35 no.1
    • /
    • pp.3-18
    • /
    • 2019
  • Urban sprawl has been criticized due to its negative effects, including the encroachment of farmland and open spaces, the increase in traffic congestion and air pollution, the decline of central city, the decrease in social capital, and the unfairness of tax burdens on infrastructure and public services. This study measures urban sprawl in the capital region of South Korea where the characteristics of urban sprawl have been known to be different from those identified in the U.S. metropolitan areas. In particular, the study examines whether the capital region has experienced the decline of the central city with an expansion of low density residential development in suburban areas. Three measurements, the sprawl index with population density, the ratio of changes in urbanized areas to changes in population, and the population density gradient, were employed to measure urban sprawl, and GIS mapping and descriptive analysis were used to examine the central city decline and the characteristics of development patterns in suburban areas. The results show that the capital region of South Korea is moving to the American style sprawled development with the decline of the central city and an increase of single detached homes in suburban areas, implying that policy makers need to develop growth management strategies to prevent urban sprawl and its negative effects that many U.S. metropolitan areas have suffered from.

Determinations of Toltrazuril and Toltrazuril Sulfone Levels in Olive Flounder Paralichthys olivaceus Samples Using Liquid Chromatography-Electrospray Ionization Tandem Mass Spectrometry (LC-MS/MS를 이용한 넙치(Paralichthys olivaceus)시료의 톨트라주릴 및 톨트라주릴 설폰 분석)

  • Hong, Do Hee;Kim, Ah Hyun;Lee, Ka Jeong;Yoon, Minchul;Son, Kwang Tae;Kim, Myoung Sug;Kim, Na Young;Jung, Sung Hee;Jo, Mi Ra
    • Korean Journal of Fisheries and Aquatic Sciences
    • /
    • v.52 no.5
    • /
    • pp.461-467
    • /
    • 2019
  • Several studies investigating the prevention and treatment of external parasites in farmed olive flounder Paralichthys olivaceus have found that the anticoccidial agent toltrazuril sulfone is an effective antiparasitic. Prior to undertaking a full-scale study, we developed analytical methods to detect the levels of toltrazuril and toltrazuril sulfone in farmed flounder samples using liquid chromatography-electrospray ionization tandem mass spectrometry (LC-MS/MS). This analysis showed that LC-MS/MS changed the mobile phase and collision energy of toltrazuril and toltrazuril sulfone. This was validated using established conditions. Sample pre-treatment for this process involved extraction with dichloromethane and purification by liquid-liquid extraction in formic acid, acetonitrile, and h-hexane, followed by determination of all compounds by LC-MS/MS. Separation was achieved within 10 min by gradient elution using a Capcell Pak C18 ($3.0{\mu}m$, $100{\times}2.0mm$) analytical column (Shiseido UG 120V) with a mixture of 0.1% (v/v) formic acid and acetonitrile. Multiple reaction monitoring was used for selective detection of toltrazuril and toltrazuril sulfone. This method yields satisfactory results for linearity, precision, and limits of quantification. Therefore, the method established in our study will serve as a basis for further research on parasite control by toltrazuril and toltrazuril sulfone.

Changes of Efficacy of Antioxidant, Antidyslipidemic, Antidiabetic and Microbiological Characteristics in Fermented and Salt-treated Fermented Codonopsis lanceolata (발효 더덕 및 소금 처리 발효 더덕의 미생물 특성과 항산화, 항비만, 항당뇨 효능 변화)

  • Seong, Eun-Hak;Lee, Myeong-Jong;Kim, Hojun;Shin, Na Rae
    • Journal of Korean Medicine for Obesity Research
    • /
    • v.18 no.2
    • /
    • pp.106-114
    • /
    • 2018
  • Objectives: We investigated about the microbial properties and changes in the efficacy of the Codonopsis lanceolata (CL) by natural fermentation. Methods: CL was fermented for four weeks in a well-ventilated place with 2.5% salt. pH, total sugar, total polyphenol, and total flavonoid were measured to determine fermentation characteristics according to fermentation period and salt treatment. Polymerase chain reaction denaturing gradient gel electrophoresis and random amplification of polymorphic DNA-polymerase chain reaction were carried out for microbial analysis during fermentation. In addition, HepG2 cell was cultured to check the lipid accumulation through oil red O staining and the glucose uptake was analyzed by measuring the 2-NBDG at C2C12 cell. Results: The pH level and the total sugar decreased with the CL fermentation. Total polyphenol and flavonoid increased after CL fermentation. It was confirmed that Leuconostoc mesenteroides were maintained continuously during fermentation. In the salt treatment CL, there was a sharp increase in Rahnella aquatilis. Lactobacillus plantarum matrix was observed in fermented CL. In addition, Lactococcus lactis, Weissella koreensis, R. aquatilis, L. plantarum, Leu. mesenteroides have been added to the salt treatment. Glucose uptake were significantly increased after fermentation with salt for four weeks. Lipid accumulation in the HepG2 cells was observed that there was difference (P<0.01) between free fatty acid group (100%) and decreased 4 weeks after fermentation (90.38%) at $800{\mu}g/mL$. Conclusions: Total polyphenol and flavonoid were increased after CL fermentation. Especially, percentage of the glucose uptake and lipid accumulation inhibition increased in CL fermentation with salt. It is expected that fermentation of salt treated CL will be more effective in diabetes and fatty liver.

Prediction of the direction of stock prices by machine learning techniques (기계학습을 활용한 주식 가격의 이동 방향 예측)

  • Kim, Yonghwan;Song, Seongjoo
    • The Korean Journal of Applied Statistics
    • /
    • v.34 no.5
    • /
    • pp.745-760
    • /
    • 2021
  • Prediction of a stock price has been a subject of interest for a long time in financial markets, and thus, many studies have been conducted in various directions. As the efficient market hypothesis introduced in the 1970s acquired supports, it came to be the majority opinion that it was impossible to predict stock prices. However, recent advances in predictive models have led to new attempts to predict the future prices. Here, we summarize past studies on the price prediction by evaluation measures, and predict the direction of stock prices of Samsung Electronics, LG Chem, and NAVER by applying various machine learning models. In addition to widely used technical indicator variables, accounting indicators such as Price Earning Ratio and Price Book-value Ratio and outputs of the hidden Markov Model are used as predictors. From the results of our analysis, we conclude that no models show significantly better accuracy and it is not possible to predict the direction of stock prices with models used. Considering that the models with extra predictors show relatively high test accuracy, we may expect the possibility of a meaningful improvement in prediction accuracy if proper variables that reflect the opinions and sentiments of investors would be utilized.

Clustering Performance Analysis of Autoencoder with Skip Connection (스킵연결이 적용된 오토인코더 모델의 클러스터링 성능 분석)

  • Jo, In-su;Kang, Yunhee;Choi, Dong-bin;Park, Young B.
    • KIPS Transactions on Software and Data Engineering
    • /
    • v.9 no.12
    • /
    • pp.403-410
    • /
    • 2020
  • In addition to the research on noise removal and super-resolution using the data restoration (Output result) function of Autoencoder, research on the performance improvement of clustering using the dimension reduction function of autoencoder are actively being conducted. The clustering function and data restoration function using Autoencoder have common points that both improve performance through the same learning. Based on these characteristics, this study conducted an experiment to see if the autoencoder model designed to have excellent data recovery performance is superior in clustering performance. Skip connection technique was used to design autoencoder with excellent data recovery performance. The output result performance and clustering performance of both autoencoder model with Skip connection and model without Skip connection were shown as graph and visual extract. The output result performance was increased, but the clustering performance was decreased. This result indicates that the neural network models such as autoencoders are not sure that each layer has learned the characteristics of the data well if the output result is good. Lastly, the performance degradation of clustering was compensated by using both latent code and skip connection. This study is a prior study to solve the Hanja Unicode problem by clustering.

An Analysis of Water Vapor Pressure to Simulate the Relative Humidity in Rural and Mountainous Regions (고해상도 상대습도 모의를 위한 농산촌 지역의 수증기압 분석)

  • Kim, Soo-ock;Hwang, Kyu-Hong;Hong, Ki-Young;Seo, Hee-Chul;Bang, Ha-Neul
    • Korean Journal of Agricultural and Forest Meteorology
    • /
    • v.22 no.4
    • /
    • pp.299-311
    • /
    • 2020
  • This paper analyzes the distribution of water vapor pressure and relative humidity in complex terrains by collecting weather observation data at 6 locations in the valley in Jungdae-ri, Ganjeon-myeon, Gurye-gun, Jeolla South Province and 14 locations in Akyang-myeon, Hadong-gun, Gyeongsang South Province, which form a single drainage basin in rural and mountainous regions. Previously estimated water vapor pressure used in the early warning system for agrometeorological hazard and actual water vapor pressure arrived at using the temperature and humidity that were measured at the highest density (1.5 m above ground) at every hour in the valley of Jungdae-ri between 19 December 2014 and 23 November 2015 and in the valley of Akyang between 15 August 2012 and 18 August 2013 were compared. The altitude-specific gradient of the observed water vapor pressure varied with different hours of the day and the difference in water vapor pressure between high and low altitudes increased in the night. The hourly variations in the water vapor pressure in the weather stations of the valley of Akyang with various topographic and ground conditions were caused by factors other than altitude. From the observed data of the study area, a coefficient that adj usts the variation in the water vapor pressure according to the specific difference in altitude and estimates it closer to the actual measured level was derived. Relative humidity was simulated as water vapor pressure estimated against the saturated water vapor pressure, thus, confirming that errors were further reduced using the derived coefficient than with the previous method that was used in the early warning system.

Numerical Study of SPGD-based Phase Control of Coherent Beam Combining under Various Turbulent Atmospheric Conditions (대기외란에 따른 SPGD 기반 결맞음 빔결합 시스템 위상제어 동작성능 분석)

  • Kim, Hansol;Na, Jeongkyun;Jeong, Yoonchan
    • Korean Journal of Optics and Photonics
    • /
    • v.31 no.6
    • /
    • pp.247-258
    • /
    • 2020
  • In this paper, based on a stochastic parallel gradient descent (SPGD) algorithm we study phase control of a coherent-beam-combining system under turbulent atmospheric conditions. Based on the statistical theory of atmospheric turbulence, we carry out the analysis of the phase and wavefront distortion of a laser beam propagating through a turbulent atmospheric medium. We also conduct numerical simulations of a coherent-beam-combining system with 7- and 19-channel laser beams distorted by atmospheric turbulence. Through numerical simulations, we characterize the phase-control characteristics and efficiency of the coherent-beam-combining system under various degrees of atmospheric turbulence. It is verified that the SPGD algorithm is capable of realizing 7-channel coherent beam combining with a beam-combining efficiency of more than 90%, even under the turbulent atmospheric conditions up to cn2 of 10-13 m-2/3. In the case of 19-channel coherent beam combining, it is shown that the same turbulent atmospheric conditions result in a drastic reduction of the beam-combining efficiency down to 60%, due to the elevated impact of the corresponding refractive-index inhomogeneity. In addition, by putting together the number of iterations of the SPGD algorithm required for phase locking under atmospheric turbulence and the time intervals of atmospheric phenomena, which typically are of the order of ㎲, it is estimated that hundreds of MHz to a few GHz of computing bandwidth of SPGD-based phase control may be required for a coherent-beam-combining system to confront such turbulent atmospheric conditions. We expect the results of this paper to be useful for quantitatively analyzing and predicting the effects of atmospheric turbulence on the SPGD-based phase-control performance of a coherent-beam-combining system.

Development of a Prediction Technique for Debris Flow Susceptibility in the Seoraksan National Park, Korea (설악산 국립공원 지역 토석류 발생가능성 평가 기법의 개발)

  • Lee, Sung-Jae;Kim, Gil Won;Jeong, Won-Ok;Kang, Won-Seok;Lee, Eun-Jai
    • Journal of Korean Society of Forest Science
    • /
    • v.110 no.1
    • /
    • pp.64-71
    • /
    • 2021
  • Recently, climate change has gradually accelerated the occurrence of landslides. Among the various effects caused by landslides,debris flow is recognized as particularly threatening because of its high speed and propagating distance. In this study, the impacts of various factors were analyzed using quantification theory(I) for the prediction of debris flow hazard soil volume in Seoraksan National Park, Korea. According to the range using the stepwise regression analysis, the order of impact factors was as follows: vertical slope (0.9676), cross slope (0.6876), altitude (0.2356), slope gradient (0.1590), and aspect (0.1364). The extent of the normalized score using the five-factor categories was 0 to 2.1864, with the median score being 1.0932. The prediction criteria for debris flow occurrence based on the normalized score were divided into four grades: class I, >1.6399; class II, 1.0932-1.6398; class III, 0.5466-1.0931; and class IV, <0.5465. Predictions of debris flow occurrence appeared to be relatively accurate (86.3%) for classes I and II. Therefore, the prediction criteria for debris flow will be useful for judging the dangerousness of slopes.

Effect of Wind Speed Profile on Wind Loads of a Fishing Boat (풍속 분포곡선이 어선의 풍하중에 미치는 영향에 관한 연구)

  • Lee, Sang-Eui
    • Journal of the Korean Society of Marine Environment & Safety
    • /
    • v.26 no.7
    • /
    • pp.922-930
    • /
    • 2020
  • Marine accidents involving fishing boats, caused by a loss of stability, have been increasing over the last decade. One of the main reasons for these accidents is a sudden wind attacks. In this regard, the wind loads acting on the ship hull need to be estimated accurately for safety assessments of the motion and maneuverability of the ship. Therefore, this study aims to develop a computational model for the inlet boundary condition and to numerically estimate the wind load acting on a fishing boat. In particular, wind loads acting on a fishing boat at the wind speed profile boundary condition were compared with the numerical results obtained under uniform wind speed. The wind loads were estimated at intervals of 15° over the range of 0° to 180°, and i.e., a total of 13 cases. Furthermore, a numerical mesh model was developed based on the results of the mesh dependency test. The numerical analysis was performed using the RANS-based commercial solver STAR-CCM+ (ver. 13.06) with the k-ω turbulent model in the steady state. The wind loads for surge, sway, and heave motions were reduced by 39.5 %, 41.6 %, and 46.1 % and roll, pitch, and yaw motions were 48.2 %, 50.6 %, and 36.5 %, respectively, as compared with the values under uniform wind speed. It was confirmed that the developed inlet boundary condition describing the wind speed gradient with respect to height features higher accuracy than the boundary condition of uniform wind speed. The insights obtained in this study can be useful for the development of a numerical computation method for ships.