• 한국경영과학회지
• /
• 제30권4호
• /
• pp.27-43
• /
• 2005

This paper presents a new hybrid data mining technique using error pattern modeling to improve classification accuracy when the data type of a target variable is binary. The proposed method increases prediction accuracy by combining two different supervised learning methods. That is, the algorithm extracts a subset of training cases that are predicted inconsistently by both methods, and models error patterns from the cases. Based on the error pattern model, the Predictions of two different methods are merged to generate final prediction. The proposed method has been tested using practical 10 data sets. The analysis results show that the performance of proposed method is superior to the existing methods such as artificial neural networks and decision tree induction.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회 2005년도 추계학술대회 및 정기총회
• /
• pp.409-413
• /
• 2005

The logical analysis of data(LAD) is an effective Boolean-logic based data mining tool. A critical step in analyzing data by LAD is the pattern generation stage where useful knowledge and hidden structural information in data is discovered in the form of patterns. A conventional method for pattern generation in LAD is based on term enumeration that renders the generation of higher degree patterns practically impossible. In this paper, we present a new optimization-based pattern generation methodology and propose two mathematical programming medels, a mixed 0-1 integer and linear programming(MILP) formulation and a well-studied set covering problem(SCP) formulation for the generation of optimal and heuristic patterns, respectively. With benchmark datasets, we demonstrate the effectiveness of our models by automatically generating with much ease patterns of high complexity that cannot be generated with the conventional approach.

• Smart Structures and Systems
• /
• 제7권2호
• /
• pp.117-132
• /
• 2011

This paper is aimed at presenting two methods on the basis of pattern search and genetic algorithms to detect and estimate damage in plates using the modal data of a damaged plate. The proposed methods determine the damages of plate structures using optimization of an objective function by pattern search and genetic algorithms. These methods have been applied to two numerical examples, namely four-fixed supported and cantilever plates with and without noise in the modal data and containing one or several damages. The obtained results clearly reveal that the proposed methods can be viewed as a powerful and reliable method for structural damage detection in plates using the modal data.

• 응용통계연구
• /
• 제18권2호
• /
• pp.371-379
• /
• 2005

본 논문에서는 피어슨 상관계수를 이용한 시간경로 유전자 발현자료의 군집분석에서 군집의 대표적인 패턴에서 벗어나는 이질적인 패턴을 보이는 시계열을 탐지하기 위한 패턴일치지수를 제안하고, 이를 마이크로어레이 실험으로부터 얻어진 혈청 시간경로 유전자 발현자료에 적용하여 유용성을 검토해 본다.

• Journal of Information Technology Applications and Management
• /
• 제15권1호
• /
• pp.225-242
• /
• 2008

PM (Inconsistency Pattern Modeling) is a hybrid supervised learning technique using the inconsistence pattern of input variables in mining data sets. The IPM tries to improve prediction accuracy by combining more than two different supervised learning methods. The previous related studies have shown that the IPM was superior to the single usage of an existing supervised learning methods such as neural networks, decision tree induction, logistic regression and so on, and it was also superior to the existing combined model methods such as Bagging, Boosting, and Stacking. The objectives of this paper is explore the characteristics of the IPM. To understand characteristics of the IPM, three experiments were performed. In these experiments, there are high performance improvements when the prediction inconsistency ratio between two different supervised learning techniques is high and the distance among supervised learning methods on MDS (Multi-Dimensional Scaling) map is long.

• 한국언어정보학회지:언어와정보
• /
• 제9권1호
• /
• pp.69-92
• /
• 2005

Pattern recognition in cognitive agents is based on (i) the uninterpreted input data (e.g. parameter values) provided by the agent's hardware devices and (ii) and interpreted patterns (e.g. templates) provided by the agent's memory. Computationally, the task consists in finding the memory data corresponding best to the input data, for any given input. Once the best fitting memory data have been found, the input is recognized by applying to it the interpretation which happens to be stored with the memorized pattern. This paper presents a fast converging procedure which starts from a few initially recognized items and then analyzes the remainder of the input by systematically checking for items shown by memory to have been related to the initial items in previous encounters. In this way, known patterns are tried first, and only when they have been exhausted, an elementary exploration of the input is commenced. Efficiency is improved further by choosing the candidate to be tested next according to frequency.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2022년도 추계학술대회
• /
• pp.210-212
• /
• 2022

제조업에서 생산 설비의 상태를 모니터링하기 위해 각종 센서를 부착하고 있으며, 이를 통해 획득된 데이터의 경우 시계열 데이터인 경우가 많다. 생산 설비의 이상 여부를 판단하기 위해서는시계열 데이터로부터 패턴을 추출하는 과정이 선행되어야 하며 다양한 방법이 연구되고 있다. 본 논문에서는 수집된 다변량 시계열 데이터로부터 패턴을 추출하기 위해 매트릭스 프로파일 알고리즘을 적용하였으며, 이를 통해 현재 CNC 머신으로부터 수집 중인 다중 센서 데이터의 패턴을 추출하였다.

• 패션비즈니스
• /
• 제16권3호
• /
• pp.78-94
• /
• 2012

This study devised and drew custom sleeve patterns by using a regression equation with the data from 7 models along with the sleeve that was slightly modified to make the general-purpose sleeve pattern. To devise a general-purpose sleeve pattern, the sleeve pattern was drawn as an object for comparison by applying the Bunka drafting system (sleeve pattern by the Bunka drafting system) to the basic upper garment. Actual sleeves, made by using the three types of patterns above, were created and tested by models. Next, 30 panel members participated in a sight test of the compatibility of the sleeves to examine the validity of the sleeve drafting method acquired using the regression equation. The test proved that the custom sleeve pattern and the general-purpose sleeve pattern were more suitable for the characteristics of arm structures. Thus, the new sleeve-cap part drafting method using the regression equation was shown to have validity. As a result, since a very significant correlation was obtained for the body measurement figures and the basic pattern of the adequate basic pattern of the sleeves, this study concludes that it is possible to come up with primary data that can be widely used by increasing the number of subjects.

• 한국컴퓨터정보학회논문지
• /
• 제3권1호
• /
• pp.95-112
• /
• 1998

역전도 알고리즘은 연관 기억장치, 음성 인식, 패턴인식, 로보틱스등 여러 응용 분야에 다양하게 사용되고 있다. 그러나 새로운 학습 패턴을 추가적으로 학습시키려면 이전에학습했던 모든 패턴과 추가되는 패턴을 갖고 처음부터 새로운 학습을 수행하여야 한다. 이는 패턴의 개수가 점차 늘어날수록 학습에 소요되는 시간이 기하 급수적으로 길어지는 결과를 초래하게 된다. 따라서 주기적으로 다량의 데이터를 추가로 학습을 할 경우에 이러한 점진적 학습은 반드시 해결해야 할 문제점으로 간주된다. 본 논문에서는 기존의 신경망 구조는 그대로 유지하면서 대표 패턴을 추출해 추가 학습을 수행하는 방법을 제안하고 제안된 기법의 효율성을 위해 기계 학습 분야의 벤치마크로 많이 사용되는 Monk's data와 Iris data에 적용해 보았다.

• 인터넷정보학회논문지
• /
• 제16권6호
• /
• pp.23-30
• /
• 2015

빅 데이터가 이슈화됨에 따라 데이터 분석의 결과를 기반으로 동작하는 많은 응용들이연구되고 왔고, 대표적인 응용들은 전자상거래 시스템의 상품 추천 서비스, 검색 엔진에서의 검색 서비스, 소셜 네트워크 서비스에서의 친구 추천 서비스 등이 있다. 본 논문은 기존의 데이터 마이닝 기법 중 데이터 집합에서 나타나는 유사한 패턴들을 마이닝하는 빈발 패턴 트리와 컴퓨터 과학의 이론에 기초한 결정트리를 결합하여 결정 빈발 트리 알고리즘을 제안한다. 이는 기존의 빈발 패턴 트리 알고리즘은 패튼 트리에서 패턴 생성에 대한 정확성은 보장되나 소셜 데이터처럼 다양한 패턴이 나타는 데이터에 대해서는 많은 수의 패턴들을 생성시켜 분석에 대한 어려움이 있어, 서브트리들과의 수렴 여부를 판단하는 모델로 변형시켜 문제를 개선한다. 또한 맵리듀스로 모델링하여 분산처리를 통한 고속 처리 알고리즘을 제시한다.