• ETRI Journal
• /
• 제32권5호
• /
• pp.676-686
• /
• 2010

Mining sequential patterns is an important research issue in data mining and knowledge discovery with broad applications. However, the existing sequential pattern mining approaches consider only binary frequency values of items in sequences and equal importance/significance values of distinct items. Therefore, they are not applicable to actually represent many real-world scenarios. In this paper, we propose a novel framework for mining high-utility sequential patterns for more real-life applicable information extraction from sequence databases with non-binary frequency values of items in sequences and different importance/significance values for distinct items. Moreover, for mining high-utility sequential patterns, we propose two new algorithms: UtilityLevel is a high-utility sequential pattern mining with a level-wise candidate generation approach, and UtilitySpan is a high-utility sequential pattern mining with a pattern growth approach. Extensive performance analyses show that our algorithms are very efficient and scalable for mining high-utility sequential patterns.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제7권2호
• /
• pp.51-62
• /
• 2018

높은 유틸리티 순차 패턴 탐사는 데이터 마이닝에서 중요한 연구 주제로 간주되고 있다. 이 주제에 대해 몇 개의 알고리즘들이 제안되었지만, 그것들은 높은 유틸리티 순차 패턴 탐사의 탐색 공간이 커지는 문제에 부딪히게 된다. 한 시퀀스의 더 엄격한 유틸리티 상한 값은 탐색 공간에서 초기에 유망하지 않은 패턴들을 더 가지치기할 수 있다. 본 논문에서 새로운 유틸리티 상한 값을 제안하는데, 그것은 한 시퀀스와 그 자손 시퀀스들의 최대 예상 유틸리티인 sequence expected utility (SEU)이다. 높은 유틸리티 순차 패턴들을 탐사하는데 필수적인 정보를 유지하기 위해 각 패턴에 대한 시퀀스 유틸리티 리스트를 새로운 자료구조로 사용한다. SEU를 활용하여 높은 유틸리티 순차 패턴들을 찾아내는 알고리즘인 High Sequence Utility List-Span (HSUL-Span)을 제안한다. 서로 다른 영역의 합성 데이터세트와 실제 데이터세트에 대한 실험 결과는 HSUL-Span이 상당히 적은 수의 후보 패턴들을 생성하고 실행 시간 면에서 다른 알고리즘들보다 우수한 것을 보여준다.

• ETRI Journal
• /
• 제43권6호
• /
• pp.1024-1037
• /
• 2021

Mining high utility itemsets (HUIs) from transaction databases considers such factors as the unit profit and quantity of purchased items. Two-phase tree-based algorithms transform a database into compressed tree structures and generate candidate patterns through a recursive pattern-growth procedure. This procedure requires a lot of memory and time to construct conditional pattern trees. To address this issue, this study employs two compressed tree structures, namely, Utility Count Tree and String Utility Tree, to enumerate valid patterns and thus promote fast utility computation. Furthermore, the study presents an algorithm called single-phase utility computation (SPUC) that leverages these two tree structures to mine HUIs in a single phase by incorporating novel pruning strategies. Experiments conducted on both real and synthetic datasets demonstrate the superior performance of SPUC compared with IHUP, UP-Growth, and UP-Growth+algorithms.

• 대한산업공학회지
• /
• 제28권1호
• /
• pp.112-127
• /
• 2002

A useful pattern is a pattern that contributes much to learning. For a classification problem those patterns near the class boundary surfaces carry more information to the classifier. For a regression problem the ones near the estimated surface carry more information. In both cases, the usefulness is defined only for those patterns either without error or with negligible error. Using only the useful patterns gives several benefits. First, computational complexity in memory and time for learning is decreased. Second, overfitting is avoided even when the learner is over-sized. Third, learning results in more stable learners. In this paper, we propose a pattern 'utility index' that measures the utility of an individual pattern. The utility index is based on the bias and variance of a pattern trained by a network ensemble. In classification, the pattern with a low bias and a high variance gets a high score. In regression, on the other hand, the one with a low bias and a low variance gets a high score. Based on the distribution of the utility index, the original training set is divided into a high-score group and a low-score group. Only the high-score group is then used for training. The proposed method is tested on synthetic and real-world benchmark datasets. The proposed approach gives a better or at least similar performance.

• 인터넷정보학회논문지
• /
• 제16권6호
• /
• pp.89-95
• /
• 2015

전통적인 빈발 패턴 마이닝은 데이터베이스로부터 사용자 정의 최소 임계치 이상의 빈도수를 가지는 유효 패턴들을 식별한다. 적절한 임계치 설정은 해당 도메인에 대한 사전 지식을 요구하므로 쉬운 작업이 아니다. 따라서 임계치 설정을 통한 마이닝 결과의 정밀한 제어 불가능으로 인해 도메인 지식을 기반으로 하지 않는 패턴 마이닝 방법이 필요하게 되었다. 상위 K 빈발 패턴 마이닝은 이러한 문제를 해결하기 위해 제안되었으며, 임계치 설정 없이 상위 K개의 중요 패턴들을 마이닝 한다. 사용자는 이를 적용함으로써 데이터베이스에 상관없이 가장 높은 빈도수의 패턴부터 K번째로 높은 빈도수의 패턴까지 찾아낼 수 있다. 비록 상위 K 빈발 패턴 마이닝이 임계치 설정 없이 상위 K개의 중요 패턴들을 마이닝 하지만, 트랜잭션 내 아이템 수량과 데이터베이스 내 서로 다른 아이템 중요도를 고려하지 못하여 많은 실세계 응용의 요구에 부합하지 못한다. 하이 유틸리티 패턴 마이닝은 아이템 중요도가 포함된 비 바이너리 데이터베이스의 특성을 고려하기 위해 제안되었으나 최소 임계치를 필요로 한다. 최근 임계치 설정 없는 하이 유틸리티 패턴 마이닝을 위한 상위 K 하이 유틸리티 패턴 마이닝이 개발되었으며, 이를 통해 사용자는 사전 지식 없이 원하는 수의 패턴을 마이닝 할 수 있다. 본 논문은 상위 K 하이 유틸리티 패턴 마이닝을 위한 알고리즘을 분석한다. 최신 알고리즘에 대한 성능분석을 통해 개선사항 및 발전 방향에 대해 고찰한다.

• 대한가정학회지
• /
• 제32권5호
• /
• pp.67-82
• /
• 1994

The major purpose of this study was to find the influencing factors in explaining information search patterns and to find if significant differences exist in search outcomes by search patterns. The data for this study were collected in a survey conducted in March of 1993. The final sample consisted of 327 respondents purchased refrigerator 340 purchased bed. The important findings of this study are as follows: First The variables related to search cost-benefit play an important role in identifying search patterns of consumers. Second search outcomes were different among four information search patterns for each of information sources. The overall search outcomes the level of purchase knowledge and of post-purchase satisfaction was relatively high for high-search and high-reliance group compared with other groups. And the results also indicate that although some consumers search less than others they still can make good purchase decision-making and can maximize their utility if they choose useful information sources selectively and use those selected information sources effectively. The findings of this study provide some implications regarding consumer education programs the consumer information providing policies and future research methods.

• Journal of information and communication convergence engineering
• /
• 제20권2호
• /
• pp.103-112
• /
• 2022

The goal of pattern mining is to identify novel patterns in a database. High utility itemset mining (HUIM) is a research direction for pattern mining. This is different from frequent itemset mining (FIM), which additionally considers the quantity and profit of the commodity. Several algorithms have been used to mine high utility itemsets (HUIs). The original BPSO algorithm lacks local search capabilities in the subsequent stage, resulting in insufficient HUIs to be mined. Compared to the transfer function used in the original PSO algorithm, the V-shaped transfer function more sufficiently reflects the probability between the velocity and position change of the particles. Considering the influence of the acceleration factor on the particle motion mode and trajectory, a nonlinear acceleration strategy was used to enhance the search ability of the particles. Experiments show that the number of mined HUIs is 73% higher than that of the original BPSO algorithm, which indicates better performance of the proposed algorithm.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2007년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.61-63
• /
• 2007

In high field (> 3 T) MR imaging, the magnetic field inhomogeneity in the target object increases due to the nonuniform electro-magnetic characteristics and relatively high Larmor frequency. Especially in the body imaging, the effect causes more serious problems resulting in locally high SAR(Specific Absorption Ratio). In this paper, we propose an optimized parallel-transmission RF coil element structure and show the utility of the coil by FDTD simulations to overcome the unwanted effects. Three types of TX coil elements are tested to maximize the efficiency and their driving patterns(amplitude and phase) optimized to have adequate field homogeneity, proper SAR level, and sufficient field strength. For the proposed coil element of 25 cm ${\times}$ 8 cm loop structure with 12 channels for a 3.0 T body coil, the 73% field non-uniformity without optimization was reduced to about 26% after optimization of driving patterns. The experimental as well as simulation results show the utility of the proposed parallel driving scheme is clinically useful for (ultra) high field MRI.

• 한국의류학회지
• /
• 제46권2호
• /
• pp.250-271
• /
• 2022

Traditional wedding dresses have had a high market demand in China in recent years. Traditional wedding dresses from the Tang dynasty occupy an important position among traditional Chinese dresses, and they are also favored by young women. This study was conducted to develop the rushan and qun patterns of traditional wedding dress styles from the Tang dynasty for women in their twenties in China. For this purpose, the rushan and qun patterns of Tang and Song dynasty dresses and modern traditional dresses were collected and analyzed. Additionally, the developed patterns were validated for suitability through appearance evaluations of virtual and real fittings. The following proportions of the developed patterns were proposed: H/3.3 for rushan length, H/33 for collar width, H/1.08 for total sleeve length, H/6 for sleeve width, H/8.5 for sleeve hem width, and H/1.55 for qun length. In addition, the developed patterns received high scores in the appearance evaluations of the virtual and real fittings. Therefore, the developed rushan and qun patterns are expected to have high utility in the current traditional wedding dress industry.

• 한국정보과학회논문지:데이타베이스
• /
• 제36권5호
• /
• pp.341-351
• /
• 2009

유틸리티 패턴 마이닝은 데이터 항목에 대한 다른 가중치를 고려할 수 있는 장점으로 인하여 비즈니스 데이터를 분석하는 환경에서 효율적으로 이용되고 있다. 그러나 기존의 빈발 패턴(Frequent Pattern) 마이닝에서의 Apriori 규칙을 그대로 적용하기 어려운 문제점으로 인하여 패턴 마이닝의 성능이 현저하게 떨어지고 있다. 본 연구는 Prefix-tree를 이용하여 지속적으로 증가하는 비즈니스 트랜잭션 데이터베이스에 대한 유틸리티 패턴 마이닝을 효과적으로 수행하기 위한 기법을 제안한다. 제안하는 기법은 Prefix-tree의 각 항목 노드에 유틸리티 값을 저장하여 FP-Growth 알고리즘에서와 같이 트리의 상향 탐색을 통하여 높은 유틸리티 패턴을 빠르게 찾아낸다. 여러 형태의 실험을 통하여 이용할 수 있는 세가지 다른 Prefix-tree 구조들 간의 성능적 특징과 패턴 탐색의 방법들을 비교하였으며 실험 결과에 따라 제안하는 기법이 기존의 기법들에 비해 많은 성능 향상을 가져올 수 있는 것을 입증하였다.