• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2021년도 제64차 하계학술대회논문집 29권2호
• /
• pp.313-315
• /
• 2021

Timely and accurate diagnosis of lung diseases using Chest X-ray images has been gained much attention from the computer vision and medical imaging communities. Although previous studies have presented the capability of deep convolutional neural networks by achieving competitive binary classification results, their models were seemingly unreliable to effectively distinguish multiple disease groups using a large number of x-ray images. In this paper, we aim to build an advanced approach, so-called Ensemble Knowledge Distillation (EKD), to significantly boost the classification accuracies, compared to traditional KD methods by distilling knowledge from a cumbersome teacher model into an ensemble of lightweight student models with parallel branches trained with ground truth labels. Therefore, learning features at different branches of the student models could enable the network to learn diverse patterns and improve the qualify of final predictions through an ensemble learning solution. Although we observed that experiments on the well-established ChestX-ray14 dataset showed the classification improvements of traditional KD compared to the base transfer learning approach, the EKD performance would be expected to potentially enhance classification accuracy and model generalization, especially in situations of the imbalanced dataset and the interdependency of 14 weakly annotated thorax diseases.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.2258-2263
• /
• 2003

This paper presents the design and development of an optimal fuzzy logic controller (FLC) for a laser tracking system. An optimal intelligent fuzzy logic controller was founded on integral criterion of the fuzzy models and three-dimensional fuzzy control. Research had been also concentrated on the methods for multivariable fuzzy models for the purposes of real-time process. Simulation results have shown remarkable tracking performance of this fuzzy PID controller.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2005년도 제36회 하계학술대회 논문집 A
• /
• pp.175-177
• /
• 2005

This paper analyzes transient behavior due to switching overvoltage in 22.9kV combined distribution systems. Computer models are consisted of distribution overhead line model, underground cable model and surge-arrester model in this paper. The computer models are made by EMTP/ATPDraw simulation and Line constants are calculated by ATP_LCC. This paper analyzes the various parameters affecting. These factors include closing angle and cable length.

• Communications for Statistical Applications and Methods
• /
• 제6권1호
• /
• pp.117-130
• /
• 1999

Nayak(1994) suggested a combined randomized response model that combined the Warner's model and greenberg et al.'s model. In this paper we extend Nayak's model to two sample case of including unknown unrelated character also propose some combined models such W-M model and G-M model that modify the Nayak's model. We suggest the efficiency conditions of our models for Nayak's model, also find the efficiency condition of G-M model for the W-M model.

• Advances in Computational Design
• /
• 제5권3호
• /
• pp.291-304
• /
• 2020

In this paper, we develop models to design the airfoil using Multilayer Feed-forward Artificial Neural Network (MFANN) and Support Vector Regression model (SVR). The aerodynamic coefficients corresponding to series of airfoil are stored in a database along with the airfoil coordinates. A neural network is created with aerodynamic coefficient as input to produce the airfoil coordinates as output. The performance of the models have been evaluated. The results show that the SVR model yields the lowest prediction error.

• International Journal of Precision Engineering and Manufacturing
• /
• 제1권2호
• /
• pp.43-47
• /
• 2000

When estimating fatigue crack growth (FCG) life of structures, the use of crack growth models and knowledge of the values of their corresponding parameters are of vital importance. Inconsistency in using models with appropriate parameters can lead to enormous errors in FCG life prediction. In this paper examples are analyzed and compared with test results to show the possible problems, Consistency checks are necessary for avoiding some pitfalls, and also necessary for verifying the correct performance and accuracy of the used computer program.

• ETRI Journal
• /
• 제32권1호
• /
• pp.160-162
• /
• 2010

For the acoustic models of embedded speech recognition systems, hidden Markov models (HMMs) are usually quantized and the original full space distributions are represented by combinations of a few quantized distribution prototypes. We propose a maximum likelihood objective function to train the quantized distribution prototypes. The experimental results show that the new training algorithm and the link structure adaptation scheme for the quantized HMMs reduce the word recognition error rate by 20.0%.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2019년도 춘계학술발표대회
• /
• pp.168-169
• /
• 2019

Trusted Execution Environment(TEE), such as Intel SGX, AMD Secure Processor and ARM TrustZone, has recently been a rising issue. Trusted Execution Environment provides a secure and independent code execution, hardware-based, environment for untrusted OS. In this paper, we show that Trusted Execution Environment's research trends on its vulnerability and attack models. We classify the previous attack models, and summarize mitigations for each TEE environment.

• Journal of Information Processing Systems
• /
• 제4권1호
• /
• pp.33-40
• /
• 2008

Traceability has been held as an important factor in testing activities as well as model-driven development. Vertical traceability affords us opportunities to improve manageability from models and test cases to a code in testing and debugging phase. This paper represents a vertical test method which connects a system test level and an integration test level in testing stage by using UML. An experiment how traceability works to effectively focus on error spots has been included by using concrete examples of tracing from models to the code.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제15권3호
• /
• pp.192-199
• /
• 2015

Heterogeneous cellular networks are gaining momentum in industry and the research community, and are attracting the attention of standard bodies such as 3GPP LTE and IEEE 802.16j, whose objectives are to increase the capacity and coverage of cellular networks. In this article, we provide an overview of expansion strategies, optimal locations of base stations with different characteristics, and radio-planning models.