• Journal of Korea Multimedia Society
• /
• v.13 no.12
• /
• pp.1786-1797
• /
• 2010

Automatic understanding of contents in document image is a very hard problem due to involvement with mathematically challenging problems originated mainly from the over-determined system induced by document segmentation process. In both academic and industrial areas, there have been incessant and various efforts to improve core parts of content retrieval technologies by the means of separating out segmentation related issues using semi-structured document, e.g., invoice,. In this paper we proposed classification models for text lines on invoice document in which text lines were clustered into the five categories in accordance with their contents: purchase order header, invoice header, summary header, surcharge header, purchase items. Our investigation was concentrated on the performance of machine learning based models in aspect of linear-discriminant-analysis (LDA) and non-LDA (logic based). In the group of LDA, na$\"{\i}$ve baysian, k-nearest neighbor, and SVM were used, in the group of non LDA, decision tree, random forest, and boost were used. We described the details of feature vector construction and the selection processes of the model and the parameter including training and validation. We also presented the experimental results of comparison on training/classification error levels for the models employed.

• KIISE Transactions on Computing Practices
• /
• v.21 no.3
• /
• pp.167-172
• /
• 2015

Machine learning is a technique for training computers to be used in classification or forecasting. Among the various types, support vector machine (SVM) is a fast and reliable machine learning mechanism. In this paper, we evaluate the stock price predictability of SVM based on financial statements, through a fundamental analysis predicting the stock price from the corporate intrinsic values. Corporate financial statements were used as the input for SVM. Based on the results, the rise or drop of the stock was predicted. The SVM results were compared with the forecasts of experts, as well as other machine learning methods such as ANN, decision tree and AdaBoost. SVM showed good predictive power while requiring less execution time than the other machine learning schemes.

• Proceedings of the KIPE Conference
• /
• 2019.11a
• /
• pp.13-15
• /
• 2019

듀얼 인버터를 가진 개방 권선형 영구자석 동기 전동기는 같은 DC Link 전압으로 모터에 더 큰 전압을 사용할 수 있게 할 수 있다. 이 때문에 DC Link와 인버터 사이의 DC/DC Boost 컨버터의 필요성을 없애줄 뿐 아니라 배터리의 전압을 낮출 수 있어 안전상의 이점 및 BMS의 요구 조건을 낮추므로 경제적 이점을 가질 수 있다. 이 시스템은 경제적 이점 외에도 모터에 고전압을 인가함으로써 기존에 비해 고속 운전 영역 확장 또는 운전 영역을 기존과 동일하게 유지한다면 인버터의 손실 감소라는 이점 또한 얻을 수 있다. 그러나 1차단 인버터와 2차단 인버터의 합성전압 차이로 인해 생기는 Zero Sequence Voltage (ZSV)로 인해 시스템의 손실을 증대시키는 Zero Sequence Current (ZSC)가 흐를 수 있다. 본 연구에서는 이를 억제하기 위한 스위칭 패턴 형성을 위한 PWM 캐리어 생성 방법에 대해 제시하고, 이 방법을 적용하여 스위칭 순서 상 센터 영벡터의 유/무에 따른 제어 영향성 분석을 추가로 진행하여 전류 리플을 줄일 수 있는 스위칭 패턴 생성 방법을 제시한다.

• The Transactions of the Korean Institute of Electrical Engineers B
• /
• v.53 no.11
• /
• pp.664-670
• /
• 2004

This paper presents a new velocity estimation strategy of a non-salient permanent magnet synchronous motor (PMSM) drive without high frequency signal injection or special PWM pattern. This approach is based on the d-axis current regulator output voltage of the drive system that has the information of rotor position error. The rotor velocity can be estimated through a rotor position tracking PI controller that controls the position error to zero. For zero and low speed operation, PI controller gains of rotor position tracking controller have a variable structure according to the estimated rotor velocity. In order to boost the bandwidth of PI controller around zero speed, a loop recovery technique is applied to the control system. The proposed method only requires the flux linkage of permanent magnet and is insensitive to the parameter estimation error and variation. The designers can easily determine the possible operating range with a desired bandwidth and perform the vector control even at low speeds. The experimental results show the satisfactory operation of the proposed sensorless algorithm under rated load conditions.

• Journal of Power Electronics
• /
• v.15 no.4
• /
• pp.1026-1034
• /
• 2015

In this paper, a unified control strategy using the current space vector modulation (CSVM) technique is proposed and applied to a bidirectional three-phase DC/AC converter. The operation of the converter changes with the direction of the power flow. In the charging mode, it works as a buck type rectifier; and during the discharging mode, it operates as a boost type inverter, which makes it suitable as an interface between high voltage AC grids and low voltage energy storage devices. This topology has the following advantages: high conversion efficiency, high power factor at the grid side, tight control of the charging current and fast transition between the charging and discharging modes. The operating principle of the mode analysis, the gate signal generation, the general control strategy and the transition from a constant current (CC) to a constant voltage (CV) in the charging mode are discussed. The proposed control strategy has been validated by simulations and experimental results obtained with a 1kW laboratory prototype using supercapacitors as an energy storage device.

• Journal of Drive and Control
• /
• v.16 no.3
• /
• pp.23-32
• /
• 2019

This study proposed a fault tolerant control algorithm for electro-hydraulic brake systems where permanent magnet synchronous motor (PMSM) drive is adopted to boost the braking pressure. To cope with motor position sensor faults in the PMSM drive, a braking pressure controller based on an open-loop speed control method for the PMSM was proposed. The magnitude of the current vector was determined from the target braking pressure, and motor rotational speed was derived from the pressure control error to build up the braking pressure. The position offset of the pump piston resulting from a leak in the hydraulic system is also compensated for using the open-loop speed control by moving the piston backward until it is blocked at the end of stroke position. The performance and stability of the proposed controller were experimentally verified. According to the results, the control algorithm can be utilized as an effective means of degraded control for electro-hydraulic brake systems in the case that a motor position sensor fault occurs.

• Journal of Information Processing Systems
• /
• v.17 no.1
• /
• pp.178-190
• /
• 2021

Alzheimer's disease (AD) is an insidious and degenerative neurological disease. It is a new topic for AD patients to use magnetic resonance imaging (MRI) and computer technology and is gradually explored at present. Preprocessing and correlation analysis on MRI data are firstly made in this paper. Then kernel principal component analysis (KPCA) is used to extract features of brain gray matter images. Finally supervised classification schemes such as AdaBoost algorithm and support vector machine algorithm are used to classify the above features. Experimental results by means of AD program Alzheimer's Disease Neuroimaging Initiative (ADNI) database which contains brain structural MRI (sMRI) of 116 AD patients, 116 patients with mild cognitive impairment, and 117 normal controls show that the proposed method can effectively assist the diagnosis and analysis of AD. Compared with principal component analysis (PCA) method, all classification results on KPCA are improved by 2%-6% among which the best result can reach 84%. It indicates that KPCA algorithm for feature extraction is more abundant and complete than PCA.

• Biomedical Science Letters
• /
• v.26 no.4
• /
• pp.249-255
• /
• 2020

Radiation therapy is one of the primary options for the treatment of malignant tumors. Even though it is an effective anti-cancer treatment, it can cause serious complications owing to radiation-induced damage to the normal tissue around the tumor. It was recently reported that normal stem cell response to the genotoxic stress of ionizing radiation can boost the therapeutic effectiveness of radiation by repairing damaged cells. Therefore, we focused on annexin A-1 (ANXA1), one of the genes induced by low-dose irradiation, and assessed whether it can protect adipose-derived stem cells (ADSCs) against oxidative stress-induced damage caused by low-dose irradiation and improve effectively cell survival. After confirming ANXA1 expression in ADSCs transfected with an ANXA1 expression vector, exposure to hydrogen peroxide (H2O2) was used to mimic cellular damage induced by a chronic oxidative environment to assess cell survival under oxidative conditions. ANXA1-transfected ADSCs demonstrated that increased viability compared with un-transfected cells and exhibited enhanced anti-oxidative properties. Taken together, these results suggest that ANXA1 could be used as a potential therapeutic target to improve the survival of stem cells after low-dose radiation treatment.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.3
• /
• pp.874-890
• /
• 2021

With the widespread deployment of the fifth-generation (5G) communication networks, various real-time applications are rapidly increasing and generating massive traffic on backhaul network environments. In this scenario, network congestion will occur when the communication and computation resources exceed the maximum available capacity, which severely degrades the network performance. To alleviate this problem, this paper proposed an intelligent resource allocation (IRA) to integrate with the extant resource adjustment (ERA) approach mainly based on the convergence of support vector machine (SVM) algorithm, software-defined networking (SDN), and mobile edge computing (MEC) paradigms. The proposed scheme acquires predictable schedules to adapt the downlink (DL) transmission towards off-peak hour intervals as a predominant priority. Accordingly, the peak hour bandwidth resources for serving real-time uplink (UL) transmission enlarge its capacity for a variety of mission-critical applications. Furthermore, to advance and boost gateway computation resources, MEC servers are implemented and integrated with the proposed scheme in this study. In the conclusive simulation results, the performance evaluation analyzes and compares the proposed scheme with the conventional approach over a variety of QoS metrics including network delay, jitter, packet drop ratio, packet delivery ratio, and throughput.

• ETRI Journal
• /
• v.45 no.3
• /
• pp.448-461
• /
• 2023

This article performs a detailed data scrutiny on a chronic kidney disease (CKD) dataset to select efficient instances and relevant features. Data relevancy is investigated using feature extraction, hybrid outlier detection, and handling of missing values. Data instances that do not influence the target are removed using data envelopment analysis to enable reduction of rows. Column reduction is achieved by ranking the attributes through feature selection methodologies, namely, extra-trees classifier, recursive feature elimination, chi-squared test, analysis of variance, and mutual information. These methodologies are ranked via Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) using weight optimization to identify the optimal features for model building from the CKD dataset to facilitate better prediction while diagnosing the severity of the disease. An efficient hybrid ensemble and novel similarity-based classifiers are built using the pruned dataset, and the results are thereafter compared with random forest, AdaBoost, naive Bayes, k-nearest neighbors, and support vector machines. The hybrid ensemble classifier yields a better prediction accuracy of 98.31% for the features selected by extra tree classifier (ETC), which is ranked as the best by TOPSIS.