• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.12
• /
• pp.108-115
• /
• 1994

Most pattern classifiers have been designed based on the ML (Maximum Likelihood) training algorithm which is simple and relatively powerful. The ML training is an efficient algorithm to individually estimate the model parameters of each class under the assumption that all class models in a classifier are statistically independent. That assumption, however, is not valid in many real situations, which degrades the performance of the classifier. In this paper, we propose a minimum-error-rate training algorithm based on the MAP (Maximum a Posteriori) approach. The algorithm regards the normalized outputs of the classifier as estimates of the a posteriori probability, and tries to maximize those estimates. According to Bayes decision theory, the proposed algorithm satisfies the condition of minimum-error-rate classificatin. We apply this algorithm to NPM (Neural Prediction Model) for speech recognition, and derive new disrminative training algorithms. Experimental results on ten Korean digits recognition have shown the reduction of 37.5% of the number of recognition errors.

• International Journal of Industrial Entomology and Biomaterials
• /
• v.1 no.1
• /
• pp.25-28
• /
• 2000

A comprehensive study on daily pupal weight loss due to histolysis and histgenesis in Bombyx mori Linn., has been undertaken. The percentag of pupal weight loss in the male pupa is higher than the female, conforming that the female pupa require more energy in the form of less pupal weight loss for ovipositional activities. The regression equation clearly shows higher percentage loss of daily pupal weight in male than female with respect to age and, this is also evident from the slope of the regression line. Moreover, analysis of the coefficient of correlation shows that the weight loss of pupa is directly correlated with the age rather than climatic parameters under which the pupa dwells.

• International Journal of Industrial Entomology and Biomaterials
• /
• v.1 no.1
• /
• pp.65-71
• /
• 2000

A study was conducted to examine the relationship between abiotic factors and the population of white-fly (Dialeuropora decempuncta) in mulberry field. The study reveals that relationship between abiotic factors and the whitefly population is very much existent like other pests in other agricultural crops. Duration and time of distinct occurrence of whitefly in mulberry is influenced by the abiotic conditions of field. Abiotic parameters of previous month are more important in influencing the intensity of the pest than the current abiotic factors. Not all the abiotic factors are equally important but factors like minimum temperature, fluctuation in temperature during the days minimum relative humidity, fluctuation in relative humidity and rainfall are the major important lactors in influencing the intensity of the pest under consideration.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.27 no.3
• /
• pp.346-354
• /
• 2024

Deep learning models require high computing power due to a substantial amount of computation. It is difficult to use them in devices with limited computing environments, such as coastal surveillance equipments. In this study, a lightweight model is constructed by analyzing the weight changes of the convolutional layers during the training process based on MobileNet and then pruning the layers that affects the model less. The performance comparison results show that the lightweight model maintains performance while reducing computational load, parameters, model size, and data processing speed. As a result of this study, an effective pruning method for constructing lightweight deep learning models and the possibility of using equipment resources efficiently through lightweight models in limited computing environments such as coastal surveillance equipments are presented.

• Journal of the Korean Society for Precision Engineering
• /
• v.24 no.6
• /
• pp.121-130
• /
• 2007

In this paper, we present the development of a new training system for equilibrium sense and postural control. This system consists of an unstable platform, a force plate, a computer, and training programs. The unstable platform provides 360 degrees of movement allowing for training in all directions. To evaluate the effects of the training system, we performed various experiments to train the ability of equilibrium sense and postural control of fifteen young healthy subjects. We measured the time a subject maintains his or her center of pressure on a target, the time a subject moves his or her center of pressure to a target, and the mean absolute deviation of the trace before and after the training. We analyzed these parameters obtained before and after the training using paried-sample T-test. The result shows that the subjects experienced distinctive enhancement in their ability of postural control through the training using our system.

• Physical Therapy Korea
• /
• v.23 no.2
• /
• pp.1-10
• /
• 2016

Background: Many studies regarding task-oriented training have recently demonstrated functional improvement in patients with post-stroke hemiparesis. The task-oriented approach is very diverse, and chronic stroke patients must have access to a sustained systematic treatment program to enhance their walking ability. Objectives: This study aimed to compare the effects of the task-oriented circuit training and treadmill training on walking function and quality of life in patients with chronic stroke. Methods: Fourteen patients with chronic stroke volunteered for this study. The subjects were randomly divided into a task-oriented circuit training group and a treadmill training group with 7 patients in each. Each training regimen was performed for 30 min a day and 3 days a week for 4 weeks. Assessment tools included the Timed Up-and-Go Test (TUGT), 10-m Walk Test, 6-min Walk Test (6MWT), and the Stroke Impact Scale (SIS). Results: The change in results of the TUGT, 6MWT, and SIS measured prior to and following the training regimens appeared to be significantly different between the two groups (p<.05). In addition, after the intervention, significant differences were found for all parameters in the task-oriented circuit training group and for the TUGT, 6MWT, and SIS in the treadmill training group (p<.05). Conclusion: The findings suggest that task-related circuit training and treadmill training may be helpful to improve walking function and quality of life of patients with post-stroke hemiparesis. Additionally, a task-related circuit training program may achieve more favorable outcomes than a treadmill program.

• Journal of Music and Human Behavior
• /
• v.12 no.2
• /
• pp.19-36
• /
• 2015

The purpose of this study was to examine the effects of rhythmic auditory stimulation (RAS) on gait parameters, with and without the presence of a melody, for adolescents with traumatic brain injury (TBI). Three adolescents with TBI received a total of ten individual RAS training sessions. At pre and posttest, spatiotemporal parameters including cadence, velocity and kinematic parameters were measured using the VICON 370 Motion Analysis System. The results showed no significant difference in gait velocity between the two conditions, thus the presence of the melody condition did not impact the outcome of RAS gait training. On the other hand, all participants showed improvement in gait function after RAS training. The cadence, velocity, stride length, and symmetry were increased and the stride time was reduced after training. The motion analysis demonstrated that the movement patterns of hip and knee joints improved, as they were more similar to normal gait, which indicates that the walkings tance became more stable. The research findings indicate that rhythm is the primary factor in mediating gait functions via RAS training. This study also supports that RAS training can effectively improve the gait function for adolescents with TBI.

• Structural Engineering and Mechanics
• /
• v.47 no.3
• /
• pp.401-419
• /
• 2013

In this study, the efficiency of adaptive neuro-fuzzy inference system (ANFIS) and genetic expression programming (GEP) in predicting the effects of infill walls on base reactions and roof drift of reinforced concrete frames were investigated. Current standards generally consider weight and fundamental period of structures in predicting base reactions and roof drift of structures by neglecting numbers of floors, bays, shear walls and infilled bays. Number of stories, number of bays in x and y directions, ratio of shear wall areas to the floor area, ratio of bays with infilled walls to total number bays and existence of open story were selected as parameters in GEP and ANFIS modeling. GEP and ANFIS have been widely used as alternative approaches to model complex systems. The effects of these parameters on base reactions and roof drift of RC frames were studied using 3D finite element method on 216 building models. Results obtained from 3D FEM models were used to in training and testing ANFIS and GEP models. In ANFIS and GEP models, number of floors, number of bays, ratio of shear walls and ratio of infilled bays were selected as input parameters, and base reactions and roof drifts were selected as output parameters. Results showed that the ANFIS and GEP models are capable of accurately predicting the base reactions and roof drifts of RC frames used in the training and testing phase of the study. The GEP model results better prediction compared to ANFIS model.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.18 no.11
• /
• pp.47-56
• /
• 2019

A deep neural network (DNN) model was proposed to predict the upset in the inertia friction welding process using a database comprising results from a series of FEM analyses. For the database, the upset length, upset beginning time, and upset completion time were extracted from the results of the FEM analyses obtained with various of axial pressure and initial rotational speed. A total of 35 training sets were constructed to train the proposed DNN with 4 hidden layers and 512 neurons in each layer, which can relate the input parameters to the welding results. The mean of the summation of squared error between the predicted results and the true results can be constrained to within 1.0e-4 after the training. Further, the network model was tested with another 10 sets of welding input parameters and results for comparison with FEM. The test showed that the relative error of DNN was within 2.8% for the prediction of upset. The results of DNN application revealed that the model could effectively provide welding results with respect to the exactness and cost for each combination of the welding input parameters.

• Physical Therapy Korea
• /
• v.19 no.3
• /
• pp.20-30
• /
• 2012

The purpose of this study was to determine the effect of ankle joint mobilization with movement (MWM) on the range of motion (ROM) in the ankle, on the muscle strength of lower extremities, and on spatiotemporal gait parameters in chronic hemiplegic patients. Fifteen subjects with chronic stroke were divided into two groups: an experimental group (8 subjects) and a control group (7 subjects). Both groups attended two or three sessions of physical therapy each week. The experimental group also attended additional MWM training sessions three times a week for five weeks. For both groups, the ROM of the ankle, the muscle strength of the lower extremities, and the spatiotemporal gait parameters in paretic limbs were evaluated before and after the training period. The results showed that the experimental group experienced more significant increases than did the control group in terms of passive (6.10%) and active (21.96%) ROM of the ankle, gait velocity (12.96%), and peak torque, of the knee flexor (81.39%), the knee extensor (24.88%), and the ankle plantar flexor (41.75%)(p<.05). These results suggest that MWM training in patients with chronic stroke may be beneficial in increasing ROM in the ankle, muscle strength in the lower extremities, and gait speed.