• Journal of information and communication convergence engineering
• /
• v.11 no.4
• /
• pp.268-273
• /
• 2013

In this paper, we propose a document classification model using Web documents as a part of the training corpus in order to resolve the imbalance of the training corpus size per category. For the purpose of retrieving the Web documents closely related to each category, the proposed document classification model calculates the matching score between word features and each category, and generates a Web search query by combining the higher-ranked word features and the category title. Then, the proposed document classification model sends each combined query to the open application programming interface of the Web search engine, and receives the snippet results retrieved from the Web search engine. Finally, the proposed document classification model adds these snippet results as Web documents to the training corpus. Experimental results show that the method that considers the balance of the training corpus size per category exhibits better performance in some categories with small training sets.

• International journal of advanced smart convergence
• /
• v.11 no.2
• /
• pp.135-141
• /
• 2022

This thesis examines a significant way to enhancing and improving the term/phenomenon of performer training system in contemporary Korean theatre. To articulate the matters, this research engages in discussing and criticizing those problematic issues that we, as an instructor/trainer, have faced with through the last decades in the field of performer training and education. Specifically, we concern with the necessity of an applicable and appropriate educational/training system where each student-actor would discover his/her own adaptability by evaluating what a specific method and approach is. This atmosphere accurately provided by an instructor/trainer can also facilitate and enhance the young students' potential possibilities and/or talent, that is, as we argue a way to accomplish each performer's true nature. To achieve the goals, we underlie the necessity of establishing and/or settling performer training program/course by means of an alternative path. The research finding shows that within the atmosphere each student could share then interrogate what a possible or ideal way is according to his/her comprehensive understandings with clearer purpose: what kind of performers would you produce, train, and/or educate.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.154.2-154
• /
• 2001

This paper presented a adaptation approach based on maximum model distance (MMD) method. This method shares the same framework as they are used for training speech recognizers with abundant training data. The MMD method could adapt to all the models with or without adaptation data. If large amount of adaptation data is available, these methods could gradually approximate the speaker-dependent ones. The approach is evaluated through the phoneme recognition task on the TIMIT corpus. On the speaker adaptation experiments, up to 65.55% phoneme error reduction is achieved. The MMD could reduce phoneme error by 16.91% even when ...

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.99.1-99
• /
• 2001

This paper presented an adaptation approach based on maximum model distance (MMD) method. This method shares the same framework as they are used for training speech recognizers with abundant training data. The MMD method could adapt to all the models with or without adaptation data. If large amount of adaptation data is available, these methods could gradually approximate the speaker-dependent ones. The approach is evaluated through the phoneme recognition task on the TIMIT corpus. On the speaker adaptation experiments, up to 65.55% phoneme error reduction is achieved. The MMD could reduce phoneme error by 16.91% even when only one adaptation utterance is used.

• Journal of the Korean Society of Physical Medicine
• /
• v.12 no.1
• /
• pp.35-42
• /
• 2017

PURPOSE: The purpose of this study was to examine the effects of treadmill gait training in an adjusted position from the functional training system on the gait and balance of chronic stroke patients. METHODS: Thirty chronic stroke patients were randomly assigned to either the experimental group, who received treadmill gait training in an adjusted position, or the control group, who received regular treadmill gait training. Both groups underwent a 30-minute comprehensive rehabilitation treatment before receiving an additional 20-minute treadmill gait training. This routine was repeated five times a week for four weeks. To measure the difference before and after training in walking and balance, patients were scored on the following: 10 m walking test (10 MWT), 6 minute walking distance (6 MWD), timed up and go test (TUG), and static standing balance test (stability index). RESULTS: While post-training scores of 10 MWT, 6 MWD, TUG, and stability index for both groups increased significantly compared with pre-training (p<.05), the experimental group showed greater improvement than the control group (p<.05). The scores of the experimental group increased significantly by 9% in the 10 MWT, 11% in 6 MWD, 13% in the TUG, 8% in the stability Index (eye opened), and 10% in the stability index (eye closed). CONCLUSION: Treadmill gait training in an adjusted position from the functional training system would be a useful gait training method to improve walking and balance of chronic stroke patients.

• Journal of The Korean Society of Integrative Medicine
• /
• v.1 no.3
• /
• pp.79-95
• /
• 2013

Purpose : This study aims to investigate the effect of the Snoezelen and Rehacom programs on improvement of attention and memory, and the effect of the Snoezelen program on stress reduction. Method : This study was targeted at 11 subjects in the Snoezelen experimental group, 11 subjects in the Rehacom group and 11 subjects in the non-experimental group. As the initial evaluation, all the subjects took electroencephalography. Then, the Snoezelen group and Rehacom group did Snoezelen training and Rehacom training, respectively total 12 times(for 20 minutes twice per week for six weeks), but no training was applied to the control group. Three weeks after the training, the interim was carried out, and four weeks after the training, the final evaluation was carried out. Results : Subjects' attention increased to $58.15{\pm}4.96$ from $43.75{\pm}4.69$ during the Snoezelen training, and increased to $49.85{\pm}1.91$ from $43.28{\pm}2.71$ during the Rehacom training, which means the Snoezelen training was more effective in improving attention(P<0.05). Subjects' memory increased to $56.14{\pm}1.26$ from $43.19{\pm}3.46$ during the Snoezelen training, and increased to $50.94{\pm}4.0$ from $43.07{\pm}2.58$ during the Rehacom training. This also implies that the Snoezelen training was more effective in improving memory(P<0.05). Conclusion : Though both of the Snoezelen training and Rehacom training improved attention and memory, the Snoezelen program was more effective, and it also influenced stress resistance and physical arousal.

• Journal of the Korean Institute of Gas
• /
• v.21 no.1
• /
• pp.34-42
• /
• 2017

We developed the safety training methods which improve the current methods by overcoming the single scenario-based one-way communication between trainee and training system. We improved the design and implementation of the safety training scenarios, which is one of the most important components of the plant safety training system for safe operation of the city gas governor. The diversity and training effects of the training scenarios can be improved by interactive training between the plant safety training system and operators. The interactive training methods were developed based on Finite State Machine model which is applicable to and based on plant safety training platform. We could see the possibility of applying this method of safety training scenario system to other domain of plant safety training system.

• Journal of Korean Physical Therapy Science
• /
• v.29 no.2
• /
• pp.1-13
• /
• 2022

Background: This research was conducted to understand balance training in trunk control, balance, and walking in stroke patients. Design: Randomized controlled trial. Methods: The subjects included 40 stroke patients, of whom 20 undertook balance training using visual information and the other 20 undertook balance training using balance boards. Using visual feedback, the balance training group used a training program within the static balanced evaluation tool, while the balance training group trained using a balance board. All subjects underwent 20 mins of neurodevelopmental treatment, and both target groups underwent 10 mins each of balance training by using either visual feedback or a balance board. The treatment period lasted a total of 4 weeks, twice a day. Trunk control before and after training was evaluated with the Trunk Impairment Scale. Balance capability was assessed by the Berg Balance Scale, Functional Reach Test, Timed Up and Go test, and Static balance measurement tool. Walking capacity was measured using gait measuring equipment, and cadence and velocity were measured. Results: Both groups showed a significant improvement in their interstitial control, balance, and gait ability after the experiments compared to before the experiments (p<0.05). The difference between the two groups was not significant. The visual feedback balance training group showed a more substantial improvement than the balance board training group. Conclusion: In this study, we found that the balance training combined with visual feedback contributes to improving trunk control, balance, and gait in patients with hemiplegia due to stroke. In addition to this, I believe that balanced training combined with visual feedback can be used as a training method when considering patients who lack interstitial control, balance, and gait ability.

• Water Engineering Research
• /
• v.1 no.2
• /
• pp.129-136
• /
• 2000

Predicting water levels is a difficult task because a lot of uncertainties are included. Therefore the neural network which is appropriate to such a problem, is introduced. One day ahead forecasting of river stage in the Bocheong River is carried out by using the neural network model. Historical water levels at Snagye gauging point which is located at the downstream of the Bocheong River and average rainfall of the Bocheong River basin are selected as training data sets. With these data sets, the training process has been done by using back propagation algorithm. Then waters levels in 1997 and 1998 are predicted with the trained algorithm. To improve the accuracy, a filtering method is introduced as predicting scheme. It is shown that predicted results are in a good agreement with observed water levels and that a filtering method can overcome the lack of training patterns.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2017.04a
• /
• pp.787-790
• /
• 2017

Data quality is the main issue in the classification problems; generally, the presence of noisy instances in the training dataset will not lead to robust classification performance. Such instances may cause the generated decision tree to suffer from over-fitting and its accuracy may decrease. Decision trees are useful, efficient, and commonly used for solving various real world classification problems in data mining. In this paper, we introduce a preprocessing technique to improve the classification accuracy rates of the C4.5 decision tree algorithm. In the proposed preprocessing method, we applied the naive Bayes classifier to remove the noisy instances from the training dataset. We applied our proposed method to a real e-commerce sales dataset to test the performance of the proposed algorithm against the existing C4.5 decision tree classifier. As the experimental results, the proposed method improved the classification accuracy by 8.5% and 14.32% using training dataset and 10-fold crossvalidation, respectively.