• 반도체디스플레이기술학회지
• /
• 제22권2호
• /
• pp.40-44
• /
• 2023

The defect detection based on deep learning requires minimal loss and high accuracy to pinpoint product defects. In this paper, we confirm the loss rate of deep learning training based on disc-shaped artificial scaffold images. It is intended to compare the performance of Cross-Entropy functions used in object detection algorithms. The model was constructed using normal, defective artificial scaffold images and category cross entropy and sparse category cross entropy. The data was repeatedly learned five times using each loss function. The average loss rate, average accuracy, final loss rate, and final accuracy according to the loss function were confirmed.

• ETRI Journal
• /
• 제46권2호
• /
• pp.205-217
• /
• 2024

We propose a bandwidth prediction approach based on deep learning. The approach is intended to accurately predict the bandwidth of various types of mobile networks. We first use a machine learning technique, namely, the gradient boosting algorithm, to recognize the connected mobile network. Second, we apply a handover detection algorithm based on network recognition to account for vertical handover that causes the bandwidth variance. Third, as the communication performance offered by 3G, 4G, and 5G networks varies, we suggest a bidirectional long short-term memory model with time distribution for bandwidth prediction per network. To increase the prediction accuracy, pretraining and fine-tuning are applied for each type of network. We use a dataset collected at University College Cork for network recognition, handover detection, and bandwidth prediction. The performance evaluation indicates that the handover detection algorithm achieves 88.5% accuracy, and the bandwidth prediction model achieves a high accuracy, with a root-mean-square error of only 2.12%.

• International Journal of Computer Science & Network Security
• /
• 제24권3호
• /
• pp.107-112
• /
• 2024

Nowadays the deadly virus famous as COVID-19 spread all over the world starts from the Wuhan China in 2019. This disease COVID-19 Virus effect millions of people in very short time. There are so many symptoms of COVID19 perhaps the Identification of a person infected with COVID-19 virus is really a difficult task. Moreover it's a challenging task to identify whether a person or individual have covid test positive or negative. We are developing a framework in which we used machine learning techniques..The proposed method uses DecisionTree, KNearestNeighbors, GaussianNB, LogisticRegression, BernoulliNB , RandomForest , Machine Learning methods as the classifier for diagnosis of covid ,however, 5-fold and 10-fold cross-validations were applied through the classification process. The experimental results showed that the best accuracy obtained from Decision Tree classifiers. The data preprocessing techniques have been applied for improving the classification performance. Recall, accuracy, precision, and F-score metrics were used to evaluate the classification performance. In future we will improve model accuracy more than we achieved now that is 93 percent by applying different techniques

• International Journal of Computer Science & Network Security
• /
• 제24권9호
• /
• pp.93-96
• /
• 2024

The segmentation, detection, and extraction of infected tumour area from magnetic resonance (MR) images are a primary concern but a tedious and time taking task performed by radiologists or clinical experts, and their accuracy depends on their experience only. So, the use of computer aided technology becomes very necessary to overcome these limitations. In this study, to improve the performance and reduce the complexity involves in the medical image segmentation process, we have investigated many algorithm methods are available in medical imaging amongst them the Threshold technique brain tumour segmentation process gives an accurate result than other methods for MR images. The proposed method compare with the K-means clustering methods, it gives a cluster of images. The experimental results of proposed technique have been evaluated and validated for performance and quality analysis on magnetic resonance brain images, based on accuracy, process time and similarity of the segmented part. The experimental results achieved more accuracy, less running time and high resolution.

• 한국기계가공학회지
• /
• 제11권3호
• /
• pp.174-181
• /
• 2012

The spindle unit is a core part in high precision machine tool. Rotation accuracy of spindle unit is needed for high dignity cutting and improving the performance of machine tool. However, there are many factors to effect to rotational error motion(rotation accuracy). This study studied how rotational error motion is variation when unbalance amount is variation. Rotation accuracy of initial spindle unit is decided depending on parts and assembly such as bearing. When it is rotation, vibration and noise is appeared depending on volume of unbalance amount, so it works to decrease unbalance amount. The purpose of the study tests that unbalance amount how much effects to initial rotation condition. The result of the study shows that accuracy of parts and assembly is highly necessary to reach high rotation accuracy and unbalance amount hardly effects to initial rotation accuracy. However, it shorten spindle's life because vibration and noise is increasing by increasing unbalance amount and we can expect situation that rotation accuracy is falling by long time operation.

• 전자공학회논문지B
• /
• 제31B권7호
• /
• pp.26-32
• /
• 1994

The Lloyd-Max algorithm is an iterative scheme for design of the minimum mean square error quantizer. It is very simple in concept and easy to program into a computer. However its convergence and accuracy are primarily dependent upon the accuracy of the initial parameter. In this paper, a new initial parameter which converges to a specific value when the number of output levels becomes large is selected. And an estimator using curve fitting techique is suggested. In addition, performance of the proposed method is shown to be superior to that of the existing methods in accuracy and convergence.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1989년도 한국자동제어학술회의논문집; Seoul, Korea; 27-28 Oct. 1989
• /
• pp.200-205
• /
• 1989

The performance of conventional robot arms is inhibited by trade-off between speed and accuracy. Because these systems measure only joint angles, in spite of slow speed, they must rely on a stiff structure in order to attain positioning accuracy. Lightweight links would allow faster motion, but their flexibility would also produce positioning errors. This research is involved with the development and evaluation of an End-point Control System whose major goal is to compensate for link deflections and thus mitigate the speed versus accuracy conflict in conventional manipulator.

• 대한전자공학회논문지
• /
• 제25권3호
• /
• pp.327-334
• /
• 1988

In this paper, the performance of PRA (pel recurdive algorithm) and BMA(block matching algorithm), which are the most well-known motion estimation techniques, is compared and the effects of the motion estimation accuracy on the motion compensated coding are described. Results of computer simulation on the real images indicate that the TSS (three step search), which is one of the BMA,is slightly better than the PRA in terms of the accuracy however, the required bit rate is 6.6-8.2 Kbps higher that of the PRA because the TSS requires a transmission of motion estimation vectors.

• 한국응급구조학회지
• /
• 제13권3호
• /
• pp.29-40
• /
• 2009

Objective : This study is an experimental study which is designed to examine the differences between knowledge and self-confidence before and after theory education(CPR PPT material) based on guidelines of CPR and emergency cardiac treatment of American Heart Association(AHA, 2005) and video self-instruction program for the general public by Korean Association of Cardiopulmonary Resuscitation(KACPR), trace CPR performance ability after CPR and AED education and investigate the accuracy of artificial respiration and chest compression, and know the difference in CPR performance abilities including AED. Methods : Subjects of this study include ground crews and staffs at M airport in G province equipped with emergency equipments for CPR according to Art. 47, Sec. 2 of Emergency Medical Law, airport police, rent-a-cops, security guard, quarantine officer, custom officer, and communication, electricity, civil engineering, facility management staff, airport fire fighting staff, air mechanic, traffic controller, and airport management team among airport facility management staffs. They were given explanation of necessity of research and 147 of 220 subjects who gave consent to this research but 73 who were absent from survey were excluded were used as subjects of this study. of 147 subjects, there were 102 men and 45 women. Results : 1) Knowledge score of CPR was $6.18{\pm}0.87$ before instruction and it was increased to $15.12{\pm}1.78$ after instruction, and there was statistically significant difference. 2) Self-confidence score in CPR was $3.16{\pm}0.96$ before instruction and it was increased to $7.05{\pm}0.75$ after instruction, and there was statistically significant difference. 3) Total average score in CPR performance ability after instruction was 7.46 out of 9, performance ability was highest in confirmation of response as 144(97.95%), follwed by request of help as 140(95.25%) and confirmation of respiration as 135(91.83%), and lowest in performing artificial respiration twice(gross elevation of chest) as 97(65.98%). Accuracy of artificial respiration(%) was $28.60{\pm}16.88$ and that of chest compression(%) was $73.10{\pm}22.16$. 4) Performance ability of AED after instruction showed proper performance in power on by 141(95.91%) and attaching pad by 135(91.83%), hand-off for analyzing rhythm showed 'accuracy' in 115(78.23%) and 'non-performance' in 32(21.77%), delivery of shock and hand-off confirmation showed 'accuracy' in 109(74.14%) and 'inaccuracy' in 38(25.86%), and beginning chest compression immediately after AED was done by 105(71.42%).

• KEPCO Journal on Electric Power and Energy
• /
• 제6권4호
• /
• pp.413-418
• /
• 2020

초기 학습 데이터의 과적합으로 인한 전력망 상태예측 모델의 성능 감소를 방지하고 예측모델의 예측 정확도 유지를 통한 계속적인 현장활용을 위해서는 기계학습 모델의 예측 정확도를 지속적으로 관리할 필요가 있다. 이를 위해, 본 논문에서는 다양한 요인에 의해 끊임없이 변화하는 전력망 상태 데이터의 특성을 고려하여 예측모델의 정확성과 신뢰성을 높이고 현장 적용 가능한 수준의 품질을 유지하기 위한 기계학습 기반 전력망 상태예측 모델의 성능 유지관리 자동화 기법을 제안한다. 제안 기법은 워크플로우 관리 기술의 적용을 통해 전력망 상태예측 모델 성능 유지관리를 위한 일련의 태스크들을 워크플로우의 형태로 모델링하고 이를 자동화하여 업무를 효율화 하였다. 또한, 기존 기술에서는 시도되지 않았던 학습데이터의 통계적 특성 변화 정도와 예측의 일반화 수준을 모두 고려한 예측모델의 성능 평가를 통해 성능 결과의 신뢰성을 확보하고 이를 통해 예측 모델의 정확도를 일정 수준으로 유지관리하고 더욱 성능이 우수한 예측모델의 신규 개발이 가능하다. 결과적으로 본 논문에서 제안하는 전력망 상태예측 모델 성능 유지관리 자동화 기법을 통해 예측모델의 성능 저하문제를 해결하여 분산자원 연계 등 외부 환경의 변화에 유연한 예측모델 관리를 통해 정확성과 신뢰성이 보장된 예측 모델의 지속적인 활용이 가능하다.