• Journal of Institute of Control, Robotics and Systems
• /
• v.10 no.12
• /
• pp.1295-1304
• /
• 2004

In this paper, the Bayesian recurrent neural network is proposed to predict time series data. A neural network predictor requests proper learning strategy to adjust the network weights, and one needs to prepare for non-linear and non-stationary evolution of network weights. The Bayesian neural network in this paper estimates not the single set of weights but the probability distributions of weights. In other words, the weights vector is set as a state vector of state space method, and its probability distributions are estimated in accordance with the particle filtering process. This approach makes it possible to obtain more exact estimation of the weights. In the aspect of network architecture, it is known that the recurrent feedback structure is superior to the feedforward structure for the problem of time series prediction. Therefore, the recurrent neural network with Bayesian inference, what we call Bayesian recurrent neural network (BRNN), is expected to show higher performance than the normal neural network. To verify the proposed method, the time series data are numerically generated and various kinds of neural network predictor are applied on it in order to be compared. As a result, feedback structure and Bayesian learning are better than feedforward structure and backpropagation learning, respectively. Consequently, it is verified that the Bayesian reccurent neural network shows better a prediction result than the common Bayesian neural network.

• Korean Journal of Applied Biomechanics
• /
• v.29 no.2
• /
• pp.105-112
• /
• 2019

Objective: The main purpose of this study was to investigate the effects of wearing an ankle weight belt while performing gait in water by focusing on the effect of using ankle weights have on the gait kinematics and the muscle activities for developing optimum training strategies. Method: A total of 10 healthy male university students were recruited for the study. Each participant was instructed to perform 3 gait conditions; normal walking over ground, walking in water chest height, and walking in water chest height while using ankle weights. All walking conditions were set at control speed of $4km/h{\pm}0.05km/h$. The depth of the swimming pool was at 1.3 m, approximately chest height. The motion capture data was recorded using 6 digital cameras and the EMG was recorded using waterproof Mini Wave. From the motion capture data, the following variables were calculated for analysis; double and single support phase (s), swing phase (s), step length (%height), step rate (m/s), ankle, knee, and hip joint angles ($^{\circ}$). From the electromyography the %RVC of the lower limb muscles medial gastrocnemius, rectus femoris, erector spinae, semitendinosus, tibialis anterior, vastus lateralis oblique was calculated. Results: The results show significant differences between the gait time, and step length between the right and left leg. Additionally, the joint angular velocities and gait velocity were significantly affected by the water resistance. As expected, the use of the ankle weights increased all of the lower leg maximum muscle activities except for the lower back muscle. Conclusion: In conclusion, the ankle weights can be shown to stimulate more muscle activity during walking in chest height water and therefore, may be useful for rehabilitation purposes.

• Journal of Korean Medical classics
• /
• v.28 no.3
• /
• pp.67-77
• /
• 2015

Objectives : This study is to verify the organ weights in Nanjing based on the weights of five viscera (五臟) in autopsy studies of modern times. Methods : Contents on organ weights from many annotations and articles on Nanjing were collected. Organ weights in autopsy studies dealt in many countries including China, India, U.S. and Korea were collected. Among the data, the average weights of liver, heart, spleen, pancreas, lung and kidney of males in the age of 18 to 60 were calculated, and the ratio of each organ was examined. Based on those results, the organ weights of Nanjing were evaluated. Results & Conclusions : There is a close correspondence between the organ weight ratios of liver(肝), heart(心), lung(肺) and kidney(腎) in Nanjing and those in autopsy studies. It proves that the organ weights in Nanjing were recorded based on an actual dissection. As a result of the analysis on autopsy studies, the average organ weights and the ratio among the organs were: liver 1416g(43.0%), heart 296g(9.0%), lung 1047g(31.8%), kidney 273g(8.3%), spleen 264g(4.5%) and pancrease 113g(3.4%). The weight of liver in Nanjing shall be 4 jin and 4 liang(4斤4兩) instead of 2 jin and 4 liang(2斤4兩) to occupy proper proportion out of other organs. It is highly possible that the weight of spleen(脾) in Nanjing is including the weight of pancrease(散膏), and the weight shall be 1 jin and 1 liang(1斤1兩) or 1 jin and 2 liang(1斤2兩) instead of 2 jin and 3 liang(2斤3兩) to occupy proper proportion out of other organs.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.6
• /
• pp.1213-1220
• /
• 2011

In this paper, a new ensemble system using dynamic weighting method with added weight information into classifiers is proposed. The weights used in the traditional ensemble system are those after the training phase. Once extracted, the weights in the traditional ensemble system remain fixed regardless of the test data set. One way to circumvent this problem in the gating networks is to update the weights dynamically by adding processes making architectural hierarchies, but it has the drawback of added processes. A simple method to update weights dynamically, without added processes, is proposed, which can be applied to the already established ensemble system without much of the architectural modification. Experiment shows that this method performs better than AdaBoost.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.2
• /
• pp.443-449
• /
• 2017

Estimation of population proportion like the distribution rate of LED TV and the prevalence of a disease are often estimated based on survey sample data. Population proportion is generally considered as a special form of population mean. In complex sampling like stratified multistage sampling with unequal probability sampling, the denominator of mean may be random variable and it is estimated like ratio estimator. In this research, we examined the estimation of distribution rate based on stratified multistage sampling, and determined some numerical outcomes using stratified random sample data with about 25% of missing observations. In the data used for this research, the survey weight was determined by deterministic way. So, the weights are not random variable, and the population distribution rate and its variance estimator can be estimated like population mean estimation. When the weights are not random variable, if one estimates the variance of proportion estimator using ratio method, then the variances may be inflated. Therefore, in estimating variance for population proportion, we need to examine the structure of data and survey design before making any decision for estimation methods.

• Journal of the Korean GEO-environmental Society
• /
• v.18 no.2
• /
• pp.21-27
• /
• 2017

In order to evaluate landslide susceptibility in Yanbian region, this study analyzed 7 factors related to landslide occurrence, such as soil, geology, land use, slope, slope aspect, fault and river by Analytic Hierarchy Process (AHP), and calculated the weights of these 7 hazard-induced factors, determined the internal weights and the relative weights between various factors. According to these weights, combining the Remote Sensing technology (RS) with Geographic Information System technology (GIS), the selected area was evaluated by using GIS raster data analysis function, then landslide susceptibility chart was mapped out. The comprehensive analysis of AHP and GIS showed that there has unstable area with the potential risk of sliding in the research area. The result of landslide susceptibility agrees well with the historical landslides, which proves the accuracy of adopted methods and hazard-induced factors.

• Asian-Australasian Journal of Animal Sciences
• /
• v.25 no.10
• /
• pp.1351-1356
• /
• 2012

The objective of this study was to compare body weight, ADG, and feed:gain ratio of antibiotic-free pigs from Yorkshire dams and sired by Yorkshire (YY), Berkshire (BY), Large Black (LBY) or Tamworth (TY) boars. All the crossbred pigs in each of three trials were raised as one group from weaning to finishing in the same deep-bedded hoop, providing a comfortable environment for the animals which allowed rooting and other natural behaviors. Birth, weaning and litter weights were measured and recorded. From approximately 50 kg to market weight (125 kg), feed intake and body weights were recorded manually (body weight) or using a FIRE (Feed Intake Recording Equipment, Osborne Industries Inc. Osborne, Kansas) system with eight individual feeding stations. Feed intake data for 106 finishing pigs between 140 and 210 d of age and the resulting weights and feed conversion ratios were analyzed by breed type. Least square means for body weights (birth, weaning and to 240 d) were estimated with Proc Mixed in SAS 9.2 for fixed effects such as crossbreed and days of age within the sire breed. The differences within fixed effects were compared using least significant differences with DIFF option. Individual birth weights and weaning weights were influenced by sire breed (p<0.05). For birth weight, BY pigs were the lightest, TY and YY pigs were the heaviest but similar to each other and LBY pigs were intermediate. For weaning weights, BY and LBY pigs were heavier than TY and YY pigs. However, litter birth and weaning weights were not influenced by sire breed, and average daily gain was also not significantly different among breed types. Tamworth sired pigs had lower overall body weight gain, and feed conversion was lower in TY and YY groups than BY and LBY groups (p<0.05), however, number of observations was somewhat limited for feed conversion and for Tamworth pigs. Overall, no convincing differences among breed types were noted for this study, but growth performance in the outdoor environment was satisfactory.

• Journal of the Korea Society of Computer and Information
• /
• v.23 no.12
• /
• pp.219-226
• /
• 2018

Collaborative filtering has been most popular approach to recommend items in online recommender systems. However, collaborative filtering is known to suffer from data sparsity problem. As a simple way to overcome this problem in literature, Jaccard index has been adopted to combine with the existing similarity measures. We analyze performance of such combination in various data environments. We also find optimal weights of factors in the combination using a genetic algorithm to formulate a similarity measure. Furthermore, optimal weights are searched for each user independently, in order to reflect each user's different rating behavior. Performance of the resulting personalized similarity measure is examined using two datasets with different data characteristics. It presents overall superiority to previous measures in terms of recommendation and prediction qualities regardless of the characteristics of the data environment.

• Survey Research
• /
• v.12 no.2
• /
• pp.89-112
• /
• 2011

This study shows the results of constructing panel data using Farm Household survey and presents some examples of empirical application. This study shows that ex post constructed panel data using repeated cross-sectional survey can be used in various dynamic analyses. This paper also shows that the well known difficult problem of longitudinal weights can be easily solved by using the existing cross-sectional weights in original cross-section data. Based on these results, we propose that the National Statistical Office not only try to construct panel data, but also construct panel data by using existing repeated cross-section data. The benefits of this approach seems to be very big in establishment survey.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.2
• /
• pp.528-544
• /
• 2015

Base Station Cooperation (BSC) has been a promising technique for combating the Inter-Cell Interference (ICI) by exchanging information through a high-speed optical fiber back-haul to increase the diversity gain. In this paper, we propose a novel pilot symbol assisted data fusion scheme for distributed Uplink BSC (UBSC) based on Differential Evolution (DE) algorithm. Furthermore, the proposed scheme exploits the pre-defined pilot symbols as the sample of transmitted symbols to constitute a sub-optimal Weight Calculation (WC) model. To circumvent the non-linear programming problem of the proposed sub-optimal model, DE algorithm is employed for searching the proper fusion weights. Compared with the existing equal weights based soft combining scheme, the proposed scheme can adaptively adjust the fusion weights according to the accuracy of cooperative information, which remains the relatively low computational complexity and back-haul traffic. Performance analysis and simulation results show that, the proposed scheme can significantly improve the system performance with the pilot settings of the existing standards.