• 한국컴퓨터정보학회논문지
• /
• 제17권2호
• /
• pp.139-147
• /
• 2012

본 논문에서는 텍스트 마이닝을 통해 생물학 문헌에서 분자 수준의 사건(event) 정보를 자동으로 추출하고, 이들 사건 정보를 기반으로 새로운 생물학 지식을 자동 추론하는 텍스트 마이닝 - 추론 통합 구조의 시스템을 다룬다. 이러한 통합 구조의 지식 발견 시스템은 미리 추출되어 데이터베이스에 등록된 정보만을 입력으로 사용하는 시스템들에 비하여 최신 정보를 보다 빨리 사용할 수 있고, 미리 정의된 형식 이외의 다양한 정보를 사용할 수 있다는 장점이 있다. 반면, 텍스트 마이닝 정보 추출 결과를 그대로 사용하기 때문에 텍스트 마이닝 모듈(module)의 성능에 따라 전체 시스템의 효용성이 크게 저하될 수도 있다는 문제가 있다. 본 논문에서는 확률 기반 필터링(filtering) 방법을 제안하여, 텍스트 마이닝 결과 중 양성 오류(false positive)를 효과적으로 제거함으로써 전체 지식 발견 시스템의 정확도 및 효용성을 높이고자 한다. 본 논문에서 제안한 확률 기반 필터링 방법은 기준(baseline) 방법으로 사용된 횟수 기반 필터링 방법보다 높은 성능을 보였다.

• ETRI Journal
• /
• 제37권2호
• /
• pp.251-261
• /
• 2015

We present a novel method for real-time automatic license plate detection in high-resolution videos. Although there have been extensive studies of license plate detection since the 1970s, the suggested approaches resulting from such studies have difficulties in processing high-resolution imagery in real-time. Herein, we propose a novel cascade structure, the fastest classifier available, by rejecting false positives most efficiently. Furthermore, we train the classifier using the core patterns of various types of license plates, improving both the computation load and the accuracy of license plate detection. To show its superiority, our approach is compared with other state-of-the-art approaches. In addition, we collected 20,000 images including license plates from real traffic scenes for comprehensive experiments. The results show that our proposed approach significantly reduces the computational load in comparison to the other state-of-the-art approaches, with comparable performance accuracy.

• 한국동물번식학회:학술대회논문집
• /
• 한국동물번식학회 2004년도 춘계학술발표대회
• /
• pp.229-229
• /
• 2004

The identification of embryo-specific genes would provide insights into early embryonic development. However, the current methods employed to identify the genes that are expressed at a specific developmental stage are labor intensive and suffer from high rates of false positives. Here we employed a new and accurate reverse transcription-polymerase chain reaction (RT-PCR) technology that involves annealing control primers (ACPs) to identify the genes that are specifically or prominently expressed in bovine early blastocysts and hatched blastocysts produced in vitro. (omitted)

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2011년도 한국컴퓨터종합학술대회논문집 Vol.38 No.1(B)
• /
• pp.79-82
• /
• 2011

무인기용 비행제어 소프트웨어는 인터럽트 핸들러에서 비결정적인 수행결과를 조래하는 경합이 발생될 수 있다. 이러한 유형의 경합을 탐지하기 위한 기존 방법은 원시 프로그램의 인터럽트 핸들러를 스레드로 변환하여 정적 경합탐지 도구를 사용하므로 프로그램 수행 시 실제 발생하지 않는 부정확한 경합(false positives)를 보고한다. 본 연구는 부정확한 경합 보고를 줄이기 위해서 원시 프로그램을 POSIX 실시간 스레브 프로그램으로 변환하고 Lockset기반 탐지기법 의해서 탐지된 공유변수를 대상으로 Happens-before 관계 분석기법을 이용하여 경합을 탐지하는 동적 경합탐지 도구를 사용한다. 제시된 방법의 실험을 위해서 Knob Assembly에 탑재되는 비행제어 소프트웨어를 대상으로 정적 경합탐지 도구와 동적 경합탐지 도구의 경합탐지 결과를 비교 분석한다.

• Journal of Communications and Networks
• /
• 제12권4호
• /
• pp.375-381
• /
• 2010

A flooding attack, such as DoS or Worm, can be easily created or even downloaded from the Internet, thus, it is one of the main threats to servers on the Internet. This paper presents an online real-time network response system, which can determine whether a LAN is suffering from a flooding attack within a very short time unit. The detection engine of the system is based on the incremental mining of fuzzy association rules from network packets, in which membership functions of fuzzy variables are optimized by a genetic algorithm. The incremental mining approach makes the system suitable for detecting, and thus, responding to an attack in real-time. This system is evaluated by 47 flooding attacks, only one of which is missed, with no false positives occurring. The proposed online system belongs to anomaly detection, not misuse detection. Moreover, a mechanism for dynamic firewall updating is embedded in the proposed system for the function of eliminating suspicious connections when necessary.

• 천문학회보
• /
• 제37권1호
• /
• pp.58.1-58.1
• /
• 2012

We present a comprehensive re-analysis of stellar photometric variability in the field of open cluster M37, using our new high-precision light curves. This dataset provides a rare opportunity to explore different types of variability between short (-minutes) and long (-one month) time-scales. To investigate the variability properties of -30,000 objects, we developed new algorithms for detecting periodic, aperiodic, and sporadic variability in their light curves. About 7.5% (2,284) of the total sample exhibits convincing variations that are induced by flares, pulsations, eclipses, starspots and, in some cases, unknown causes. The benefits of our new photometry and analysis package are evident. The discovery rate of new variables is increased by 63% in comparison with the existing catalog of variables, and 51 previously identified variables were found to be false positives resulting from time-dependent systematic effects. Based on extended and improved catalog of variables, we will review the basic properties (e.g., periodicity, amplitude, type) of the variability and how different they are for different spectral types and for cluster memberships.

• 한국컴퓨터정보학회논문지
• /
• 제22권3호
• /
• pp.81-88
• /
• 2017

In this paper, we classified the weaknesses of C/C++ programs listed in CWE based on the diagnostic information produced at each stage of program compilation. Our classification identifies which stages should be responsible for analyzing the weaknesses. We also present algorithmic frameworks for detecting typical weaknesses belonging to the classes to demonstrate validness of our scheme. For the weaknesses that cannot be analyzed by using the diagnostic information, we separated them as a group that are often detectable by the analyses that simulate program execution, for instance, symbolic execution and abstract interpretation. We expect that classification of weaknesses, and diagnostic information accordingly, would contribute to systematic development of static analyzers that minimizes false positives and negatives.

• Asian Pacific Journal of Cancer Prevention
• /
• 제15권2호
• /
• pp.569-575
• /
• 2014

Although mammography is the primary imaging modality for the breast, it has its limitations especially with dense breast parenchyma. Breast magnetic resonance imaging (MRI) has evolved into an important adjunctive tool as it is currently the most sensitive technique for breast cancer detection. Despite this high sensitivity, overlap in the appearances of some benign and malignant breast lesions results in additional unnecessary intervention with negative results. These false positives, in addition to high cost and limited availability, necessitate establishing proper indications for breast MRI. The literature was here reviewed for recent clinical trials, meta-analyses and review papers which have studied this important subject. PubMed; the US national library of medicine, was utilized to review the literature in the last twenty years. Using the obtained information, current uses of breast MRI are discussed in this paper to determine the indications which are relevant to clinical practice.

• Journal of Communications and Networks
• /
• 제13권2호
• /
• pp.125-131
• /
• 2011

Rapid strides are being made to bring to reality the technology of wearable sensors for monitoring patients' physiological data.We study the problem of automatically detecting anomalies in themeasured blood glucose levels. The normal daily measurements of the patient are used to train a hidden Markov model (HMM). The structure of the HMM-its states and output symbols-are selected to accurately model the typical transitions in blood glucose levels throughout a 24-hour period. The learning of the HMM is done using historic data of normal measurements. The HMM can then be used to detect anomalies in blood glucose levels being measured, if the inferred likelihood of the observed data is low in the world described by the HMM. Our simulation results show that our technique is accurate in detecting anomalies in glucose levels and is robust (i.e., no false positives) in the presence of reasonable changes in the patient's daily routine.

• International journal of advanced smart convergence
• /
• 제8권2호
• /
• pp.116-125
• /
• 2019

The fire should extinguish as soon as possible because it causes economic loss and loses precious life. In this study, we propose a new atypical fire and smoke detection algorithm using deep learning and color histogram of fire and smoke. First, input frame images obtain from the ONVIF surveillance camera mounted in factory search motion candidate frame by motion detection algorithm and mean square error (MSE). Second deep learning (Faster R-CNN) is used to extract the fire and smoke candidate area of motion frame. Third, we apply a novel algorithm to detect the fire and smoke using color histogram algorithm with local area motion, similarity, and MSE. In this study, we developed a novel fire and smoke detection algorithm applied the local motion and color histogram method. Experimental results show that the surveillance camera with the proposed algorithm showed good fire and smoke detection results with very few false positives.