• Explosives and Blasting
• /
• v.38 no.4
• /
• pp.16-25
• /
• 2020

Multi-setting smart-investigation of the ground and large uncharged hole boring (MSP) method to reduce the blast-induced vibration in a tunnel excavation is carried out over 50m of long-distance boring in a horizontal direction and thus has been accompanied by deviations in boring alignment because of the heavy and one-directional rotation of the rod. Therefore, the deviation has been adjusted through the boring machine's variable setting rely on the previous construction records and expert's experience. However, the geological characteristics, machine conditions, and inexperienced workers have caused significant deviation from the target alignment. The excessive deviation from the boring target may cause a delay in the construction schedule and economic losses. A deep learning-based prediction model has been developed to discover an ideal initial setting of the MSP machine. Dropout, early stopping, pre-training techniques have been employed to prevent overfitting in the training phase and, significantly improved the prediction results. These results showed the high possibility of developing the model to suggest the boring machine's optimum initial setting. We expect that optimized setting guidelines can be further developed through the continuous addition of the data and the additional consideration of the other factors.

• Smart Media Journal
• /
• v.7 no.4
• /
• pp.17-23
• /
• 2018

We proposed a LSTM-based RNN model that can effectively perform the automatic spacing characteristics. For those long or noisy sentences which are known to be difficult to handle within Neural Network Learning, we defined a proper input data format and decoding data format, and added dropout, bidirectional multi-layer LSTM, layer normalization, and attention mechanism to improve the performance. Despite of the fact that Sejong corpus contains some spacing errors, a noise-robust learning model developed in this study with no overfitting through a dropout method helped training and returned meaningful results of Korean word spacing and its patterns. The experimental results showed that the performance of LSTM sequence-to-sequence model is 0.94 in F1-measure, which is better than the rule-based deep-learning method of GRU-CRF.

• PNF and Movement
• /
• v.21 no.2
• /
• pp.255-263
• /
• 2023

Purpose: This study aims to investigate the immediate effects of electromyography (EMG) biofeedback training of the gluteus medius on dynamic balance during single leg squats in healthy individuals. Methods: The sample size in this study was estimated using the G-power program at an effect size of 0.4, a significance level (α) of 0.05, and a testing power of 0.90. In addition, as a result of considering the 10% dropout rate, this study recruited 21 healthy individuals (8 males and 13 females). All subjects measured the Y-balance test-lower quarter (YBT-LQ) and limits of stability (LOS) before and after a single leg squat (SLS) and SLS with EMG biofeedback training of the gluteus medius (SLSEB). They were trained for 10 minutes for each exercise, and two dynamic balance tests were performed three times. Results: There was a significant difference in the YBT-LQ score between the two exercises (p < 0.05). In the YBT-LQ score, there was a significant difference before and after SLS and SLSEB (p < 0.05). SLSEB showed a significantly higher YBT-LQ score than SLS (p < 0.05). There was a significant difference in LOM between the two exercises (p < 0.05). However, there was no significant difference between the two exercises. Conclusion: A single-leg squat with EMG biofeedback exercises is an effective method to improve dynamic balance, such as the YBT-LQ.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.05a
• /
• pp.408-410
• /
• 2021

In the NLP field, the pre-training model BERT launched by the Google team in 2018 has shown amazing results in various tasks in the NLP field. Subsequently, many variant models have been derived based on the original BERT, such as RoBERTa, ERNIEBERT and so on. In this paper, the WWMBERT (Whole Word Masking BERT) model suitable for Chinese text tasks was used as the baseline model of our experiment. The experiment is mainly for "Text-level Chinese text classification tasks" are improved, which mainly combines Tapt (Task-Adaptive Pretraining) and "Multi-Sample Dropout method" to improve the model, and compare the experimental results, experimental data sets and model scoring standards Both are consistent with the official WWMBERT model using Accuracy as the scoring standard. The official WWMBERT model uses the maximum and average values of multiple experimental results as the experimental scores. The development set was 97.70% (97.50%) on the "text-level Chinese text classification task". and 97.70% (97.50%) of the test set. After comparing the results of the experiments in this paper, the development set increased by 0.35% (0.5%) and the test set increased by 0.31% (0.48%). The original baseline model has been significantly improved.

• The Korean Journal of Food And Nutrition
• /
• v.37 no.4
• /
• pp.197-210
• /
• 2024

This study aims to develop a deep learning model to monitor rice serving amounts in institutional foodservice, enhancing personalized nutrition management. The goal is to identify the best convolutional neural network (CNN) for detecting rice quantities on serving trays, addressing balanced dietary intake challenges. Both a vanilla CNN and 12 pre-trained CNNs were tested, using features extracted from images of varying rice quantities on white trays. Configurations included optimizers, image generation, dropout, feature extraction, and fine-tuning, with top-1 validation accuracy as the evaluation metric. The vanilla CNN achieved 60% top-1 validation accuracy, while pre-trained CNNs significantly improved performance, reaching up to 90% accuracy. MobileNetV2, suitable for mobile devices, achieved a minimum 76% accuracy. These results suggest the model can effectively monitor rice servings, with potential for improvement through ongoing data collection and training. This development represents a significant advancement in personalized nutrition management, with high validation accuracy indicating its potential utility in dietary management. Continuous improvement based on expanding datasets promises enhanced precision and reliability, contributing to better health outcomes.

• CELLMED
• /
• v.3 no.2
• /
• pp.13.1-13.6
• /
• 2013

The objective of the current study was to find out the effect of Integrated Amrita Meditation Technique (IAM) on blood pressure (BP), heart rate (HR), respiratory rate (RR) and IgA. One hundred and fifty subjects were randomized into three groups IAM, Progressive Muscle Relaxation (PMR) and Control. Baseline values were collected before giving the training for all the subjects and the IAM and PMR groups were given training in the respective techniques. BP, HR, RR and IgA were recorded manually at 0 h, 48 h, 2 months and 8 months after the first visit. HR was found to be reduced in the IAM group 48 h onwards and the fall sustained till 8 months (p < 0.05). IAM group showed significant drop when compared to the PMR group and control group in all the subsequent visits (p < 0.05). RR decreased significantly in the IAM group in the third and fourth visits (p < 0.05). RR of IAM showed significant decrease when compared to PMR and control from the third visit onwards. IgA showed significant increase in comparison with PMR and control in the third and fourth visits. BP did not show any difference in any of the visits. There was subject dropout from randomization to completion of the study, in all the three groups. The significant decrease in HR and RR and increase in IgA in the IAM group when compared to the PMR and control group shows the efficacy of the technique in reducing the physiological stress indicators for up to 8 months.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.49 no.11
• /
• pp.901-907
• /
• 2021

The present study introduces an artificial neural network (ANN) that can predict the missile aerodynamic coefficients for various missile nose shapes and flow conditions such as Mach number and angle of attack. A semi-empirical missile aerodynamics code is utilized to generate a dataset comprised of the geometric description of the nose section of the missiles, flow conditions, and aerodynamic coefficients. Data normalization is performed during the data preprocessing step to improve the performance of the ANN. Dropout is used during the training phase to prevent overfitting. For the missile nose shape and flow conditions not included in the training dataset, the aerodynamic coefficients are predicted through ANN to verify the performance of the ANN. The result shows that not only the ANN predictions are very similar to the aerodynamic coefficients produced by the semi-empirical missile aerodynamics code, but also ANN can predict missile aerodynamic coefficients for the untrained nose section of the missile and flow conditions.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.5
• /
• pp.395-401
• /
• 2023

Sound Event Detection (SED) aims to identify not only sound category but also time interval for target sounds in an audio waveform. It is a critical technique in field of acoustic surveillance system and monitoring system. Recently, various models have introduced through Detection and Classification of Acoustic Scenes and Events (DCASE) Task 4. This paper explored how to design optimal parameters of DenseNet based model, which has led to outstanding performance in other recognition system. In experiment, DenseRNN as an SED model consists of DensNet-BC and bi-directional Gated Recurrent Units (GRU). This model is trained with Mean teacher model. With an event-based f-score, evaluation is performed depending on parameters, related to model architecture as well as model training, under the assessment protocol of DCASE task4. Experimental result shows that the performance goes up and has been saturated to near the best. Also, DenseRNN would be trained more effectively without dropout technique.

• Computers and Concrete
• /
• v.30 no.5
• /
• pp.301-310
• /
• 2022

Cement-based sensors have been widely used as structural health monitoring systems, however, their long-term sensing performance have not actively investigated. In this study, a deep learning-based methodology is adopted to predict the long-term piezoresistive properties of cement-based sensors. Samples with different multi-walled carbon nanotube contents (0.1, 0.3, and 0.5 wt.%) are fabricated, and piezoresistive tests are conducted over 10,000 loading cycles to obtain the training data. Time-dependent degradation is predicted using a modified long short-term memory (LSTM) model. The effects of different model variables including the amount of training data, number of epochs, and dropout ratio on the accuracy of predictions are analyzed. Finally, the effectiveness of the proposed approach is evaluated by comparing the predictions for long-term piezoresistive sensing performance with untrained experimental data. A sensitivity of 6% is experimentally examined in the sample containing 0.1 wt.% of MWCNTs, and predictions with accuracy up to 98% are found using the proposed LSTM model. Based on the experimental results, the proposed model is expected to be applied in the structural health monitoring systems to predict their long-term piezoresistice sensing performances during their service life.

• Journal of Engineering Education Research
• /
• v.27 no.1
• /
• pp.3-12
• /
• 2024

The degree-linked programs for employees, operated by joint training centers in specialized universities that have implemented work-study integrated programs, are educational programs that require an annual government budget of around 80 billion KRW. However, the 70+ universities running these programs face issues such as a decline in academic achievement and an increase in dropout rates among students. In this paper, I conducted multiple regression analysis based on observed and measured information to examine whether the participating students in these programs are achieving an appropriate level of academic performance and to identify the factors that universities need to invest in to achieve that level. To do this, I hypothesized a causal relationship between the university's input factors and students' academic achievement, and used the SPSS program to analyze the statistical data, confirming the validity of the hypothesis. The collected data for the study were obtained through a survey developed using a Likert 4-point scale, which quantified the distribution of grades among students enrolled in IT-related departments offering the degree-linked programs for employees and the emotional contact efforts made by the universities to motivate them for academic success. Particularly, through the results of multiple regression analysis, it was confirmed that these input factors, unlike those for students in general education programs, require more personalized and frequent interactions.