• Investigative Magnetic Resonance Imaging
• /
• v.25 no.1
• /
• pp.10-22
• /
• 2021

Purpose: To overcome the difficulty in building a large data set with a high-quality in medical imaging, a concept of 'blended-transfer learning' (BTL) using a combination of both source data and target data is proposed for the target task. Materials and Methods: Source and target tasks were defined as training of the source and target networks to reconstruct cardiac CINE images from undersampled data, respectively. In transfer learning (TL), the entire neural network (NN) or some parts of the NN after conducting a source task using an open data set was adopted in the target network as the initial network to improve the learning speed and the performance of the target task. Using BTL, an NN effectively learned the target data while preserving knowledge from the source data to the maximum extent possible. The ratio of the source data to the target data was reduced stepwise from 1 in the initial stage to 0 in the final stage. Results: NN that performed BTL showed an improved performance compared to those that performed TL or standalone learning (SL). Generalization of NN was also better achieved. The learning curve was evaluated using normalized mean square error (NMSE) of reconstructed images for both target data and source data. BTL reduced the learning time by 1.25 to 100 times and provided better image quality. Its NMSE was 3% to 8% lower than with SL. Conclusion: The NN that performed the proposed BTL showed the best performance in terms of learning speed and learning curve. It also showed the highest reconstructed-image quality with the lowest NMSE for the test data set. Thus, BTL is an effective way of learning for NNs in the medical-imaging domain where both quality and quantity of data are always limited.

• Journal of Korea Multimedia Society
• /
• v.3 no.4
• /
• pp.416-423
• /
• 2000

The existing self-organizing feature map of Kohonen has weakpoint that need too much input patterns in order to converse into the learning rate and equilibrium state when it trains. Making up for the current weak point, B.Bavarian suggested the method of that distributed the learning rate such as Gaussian function. However, this method has also a disadvantage which can not achieve the right self-organizing. In this paper, we proposed the method of improving the convergence speed and the convergence rate of self-organizing feature map converting the Gaussian function into dynamic approximate curve used in when trains the self-organizing feature map.

• Knee surgery & related research
• /
• v.30 no.4
• /
• pp.303-310
• /
• 2018

Purpose: The purpose of this study was to investigate whether surgical experience could improve surgical competency in medial open wedge high tibial osteotomy (MOWHTO). Materials and Methods: One hundred consecutive cases of MOWHTO were performed with preoperative planning using the Miniaci method. Surgical errors were defined as under- or overcorrection, excessive posterior slope change, or the presence of a lateral hinge fracture. Each of these treatment failures was separately evaluated using the cumulative summation test for learning curve (LC-CUSUM). Results: The LC-CUSUM showed competency in prevention of undercorrection, excessive posterior slope change, and lateral hinge fracture after 27, 47, and 42 procedures, respectively. However, the LC-CUSUM did not signal achievement of competency in prevention of overcorrection after 100 procedures. Furthermore, the failure rate for overcorrection showed an increasing tendency as surgical experience increased. Conclusions: Surgical experience may improve the surgeon's competency in prevention of undercorrection, excessive posterior slope change, and lateral hinge fracture. However, it may not help reduce the incidence of overcorrection even after performance of 100 cases of MOWHTO over a period of 6 years.

• Geomechanics and Engineering
• /
• v.25 no.1
• /
• pp.1-16
• /
• 2021

Machine learning models have been widely used for landslide susceptibility assessment (LSA) in recent years. The large number of inputs or conditioning factors for these models, however, can reduce the computation efficiency and increase the difficulty in collecting data. Feature selection is a good tool to address this problem by selecting the most important features among all factors to reduce the size of the input variables. However, two important questions need to be solved: (1) how do feature selection methods affect the performance of machine learning models? and (2) which feature selection method is the most suitable for a given machine learning model? This paper aims to address these two questions by comparing the predictive performance of 13 feature selection-based machine learning (FS-ML) models and 5 ordinary machine learning models on LSA. First, five commonly used machine learning models (i.e., logistic regression, support vector machine, artificial neural network, Gaussian process and random forest) and six typical feature selection methods in the literature are adopted to constitute the proposed models. Then, fifteen conditioning factors are chosen as input variables and 1,017 landslides are used as recorded data. Next, feature selection methods are used to obtain the importance of the conditioning factors to create feature subsets, based on which 13 FS-ML models are constructed. For each of the machine learning models, a best optimized FS-ML model is selected according to the area under curve value. Finally, five optimal FS-ML models are obtained and applied to the LSA of the studied area. The predictive abilities of the FS-ML models on LSA are verified and compared through the receive operating characteristic curve and statistical indicators such as sensitivity, specificity and accuracy. The results showed that different feature selection methods have different effects on the performance of LSA machine learning models. FS-ML models generally outperform the ordinary machine learning models. The best FS-ML model is the recursive feature elimination (RFE) optimized RF, and RFE is an optimal method for feature selection.

• Journal of Gastric Cancer
• /
• v.10 no.4
• /
• pp.247-253
• /
• 2010

Purpose: Most stomach surgeons have been educated sufficiently in conventional open distal gastrectomy (ODG) but insufficiently in laparoscopy-assisted distal gastrectomy (LADG). We compared learning curves and clinical outcomes between ODG and LADG by a single surgeon who had sufficient education of ODG and insufficient education of LADG. Materials and Methods: ODG (90 patients, January through September, 2004) and LADG groups (90 patients, June 2006 to June 2007) were compared. The learning curve was assessed with the mean number of retrieved lymph nodes, operation time, and postoperative morbidity/mortality. Results: Mean operation time was 168.3 minutes for ODG and 183.6 minutes for LADG. The mean number of retrieved lymph nodes was 37.9. Up to about the 20th to 25th cases, the slope decrease in the learning curve for LADG was more apparent than for ODG, although they both reached plateaus after the 50th cases. The mean number of retrieved lymph nodes reached the overall mean after the 30th and 40th cases for ODG and LADG, respectively. For ODG, complications were evenly distributed throughout the subgroups, whereas for LADG, complications occurred in 10 (33.3%) of the first 30 cases. Conclusions: Compared with conventional ODG, LADG is feasible, in particular for a surgeon who has had much experience with conventional ODG, although LADG required more operative time, slightly more time to get adequately retrieved lymph nodes and more complications. However, there were more minor problems in the first 30 LADG than ODG cases. The unfavorable results for LADG can be overcome easily through an adequate training program for LADG.

• The Journal of Korean Medicine
• /
• v.43 no.2
• /
• pp.27-41
• /
• 2022

Objectives: There should be an objective analysis on the academic competency for incorporating Computer-based Test (CBT) in the education of traditional Korean medicine (TKM). However, the Item Response Theory (IRT) for analyzing latent competency has not been introduced for its difficulty in calculation, interpretation and utilization. Methods: The current study analyzed responses of 390 students of 8 years to the herbology test with 14 items by utilizing Rasch model, and the characteristics of test and items were evaluated by using characteristic curve, information curve, difficulty, academic competency, and test score. The academic competency of the students across gender and years were presented with scale characteristic curve, Kernel density map, and Wright map, and examined based on T-test and ANOVA. Results: The estimated item, test, and ability parameters based on Rasch model provided reliable information on academic competency, and organized insights on students, test and items not available with test score calculated by the summation of item scores. The test showed acceptable validity for analyzing academic competency, but some of items revealed difficulty parameters to be modified with Wright map. The gender difference was not distinctive, however the differences between test years were obvious with Kernel density map. Conclusion: The current study analyzed the responses in the herbology test for measuring academic competency in the education of TKM using Rasch model, and structured analysis for competency-based Teaching in the e-learning era was suggested. It would provide the foundation for the learning analytics essential for self-directed learning and competency adaptive learning in TKM.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.805-807
• /
• 2021

In-hospital cardiac arrest is a significant problem for medical systems. Although the traditional early warning systems have been widely applied, they still contain many drawbacks, such as the high false warning rate and low sensitivity. This paper proposed a strategy that involves a deep learning approach based on a novel interpretable deep tabular data learning architecture, named TabNet, for the Rapid Response System. This study has been processed and validated on a dataset collected from two hospitals of Chonnam National University, Korea, in over 10 years. The learning metrics used for the experiment are the area under the receiver operating characteristic curve score (AUROC) and the area under the precision-recall curve score (AUPRC). The experiment on a large real-time dataset shows that our method improves compared to other machine learning-based approaches.

• Bulletin of the Korean Mathematical Society
• /
• v.55 no.2
• /
• pp.611-624
• /
• 2018

In this paper, we study Weierstrass semigroups of ramification points on double covers of plane curves of degree 6. We determine all the Weierstrass semigroups when the genus of the covering curve is greater than 29 and the ramification point is on a total flex.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.20 no.43
• /
• pp.175-195
• /
• 1997

The learning curve has an important effect the growth of corporation. But, in Korea, the study and inference on the learning rate of each industry are unprepared, and so, Korean industires have difficult in productivity and cost. At this point, this study infers the learning rate of the oil industries and investigates the productivity and growth of them. In conclusion, this study presents the direction of the oil industries' development. With the intention of this objects, this study seizes the status which is concerned the total quantity, the operating rate, the plant capacity, the indicators concerning productivity, the investment of R & D and the scales, and then, infers and verifies the relevancy in connection with the learning rate. In the oil industry, the average rate of learning is 65.96% from 1982 to 1994 which the total quantity and the average operation time are used to infer the rate. To observe the low rate within a same period of time, this study takes the consequences that the learning rate is almost indentical with them each year. This steady state is caused by a difference between the employee and the decision maker about the acquirement and assimiliated of technology. When the high-quality technologies posses the environment to applicate in the scene of labor with them, this technology applies to the productivities. As the learning rate increases, the productivity has more effectiveness. The result of analysis about the effectiveness of the learning rate follows that the R & D unfoldes to exist and does not contribute to the growth of the oil industry. To analyze the variables of the growth, such as the learning rate, the investement of R & D, the operating rate and the gross value added to property, plant and equipment, the model is established and examined. The business strategy in the oil industry must be developed to achive the internal growth as well as the external.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.2
• /
• pp.199-209
• /
• 2022

Typical anomaly detection algorithms are trained by using prior data. Thus the batch learning based algorithms cause inevitable performance degradation when characteristics of newly incoming normal data change over time. We propose an online anomaly detection algorithm which can consider the gradual characteristic changes of incoming normal data. The proposed algorithm based on one-class classification model includes both offline and online learning procedures. In offline learning procedure, the algorithm learns the prior data to be close to centroid of the latent space and then updates the centroid of the latent space incrementally by new incoming data. In the online learning, the algorithm continues learning by using the updated centroid. Through experiments using public underwater acoustic data, the proposed online anomaly detection algorithm takes only approximately 2 % additional learning time for the incremental centroid update and learning. Nevertheless, the proposed algorithm shows 19.10 % improvement in Area Under the receiver operating characteristic Curve (AUC) performance compared to the offline learning model when new incoming normal data comes.