• 스마트미디어저널
• /
• 제10권3호
• /
• pp.23-30
• /
• 2021

제조 시계열 데이터 클러스터링 기법은 제조 대용량 데이터 기반 군집화를 통한 설비 및 공정 이상 탐지 분류를 위한 중요한 솔루션이지만 기존 정적 데이터 대상 클러스터링 기법을 시계열 데이터에 적용함에 있어 낮은 정확도를 가지는 단점이 있다. 본 논문에서는 진화 연산 기반 시계열 군집 분석 접근 방식을 제시하여 기존 클러스터링 기술에 대한 정합성 향상하고자 한다. 이를 위하여 먼저 제조 공정 결과 이미지 형상을 선형 스캐닝을 활용하여 1차원 시계열 데이터로 변환하고 해당 변환 데이터 대상으로 Pearson 거리 매트릭을 기반으로 계층적 군집 분석 및 분할 군집 분석에 대한 최적 하위클러스터를 도출한다. 해당 최적 하위클러스터 대상 유전 알고리즘을 활용하여 유사도가 최소화되는 최적의 군집 조합을 도출한다. 그리고 실제 제조 과정 이미지 대상으로 기존 클러스터링 기법과 성능 비교를 통하여 제안된 클러스터링 기법의 성능 우수성을 검증한다.

• 한국컴퓨터정보학회논문지
• /
• 제29권5호
• /
• pp.155-164
• /
• 2024

본 연구는 ESG 등급이 제공되지 않는 국내 공공기관의 ESG 등급을 추정하는 비지도 학습 기반 군집모형을 제안한다. 이를 위해, 스펙트럼 군집과 k-means 군집에서 최적의 클러스터 수를 비교했고, 그 결과의 신뢰성을 보장하기 위해 성능지표인 Davies-Bouldin Index (DBI)를 계산했다. 결과적으로, 스펙트럼 군집과 k-means 군집에서 각각 0.734 및 1.715의 DBI 값을 산출했는데, 이는 값이 작을수록 우수한 성능을 의미하므로 스펙트럼 군집의 우수성을 확인하였다. 게다가, T-검정 및 ANOVA를 이용하여 ESG 비재무 데이터 간 통계적으로 유의미한 차이를 밝혀내고, 상관계수를 이용하여 ESG 항목 간 상관관계를 확인했다. 본 연구는 이러한 결과를 바탕으로 기존 ESG 등급 없이 공공기관별 ESG 성과 순위를 추정할 가능성을 제시한다. 이는 최적의 클러스터 수를 계산한 다음, 각 클러스터 내 ESG 데이터의 평균 총합을 결정함으로써 달성된다. 따라서, 제안된 모델은 다양한 국내 공공기관의 ESG 등급을 평가하는 근거로 활용될 수 있고, 국내 지속가능경영 실천과 성과관리에 유용할 것으로 기대된다.

• The Journal of Asian Finance, Economics and Business
• /
• 제1권2호
• /
• pp.31-37
• /
• 2014

The portfolio selection is one of the most important and vital decisions that a real or legal person, who invests in stock market, should make. The main purpose of this article is the determination of the optimal portfolio with regard to relations among stock returns of companies which are active in Tehran's stock market. For achieving this goal, weekly statistics of company's stocks since Farvardin 1389 until Esfand 1390, has been used. For analyzing statistics and information and examination of stocks of companies which has change in returns, factors analysis approach and clustering analysis has been used (FC approach). With using multivariate analysis and with the aim of reducing the unsystematic risk, a financial portfoliois formed. At last but not least, results of choosing the optimal portfolio rather than randomly choosing a portfolio are given.

• 산업기술연구
• /
• 제29권B호
• /
• pp.9-15
• /
• 2009

In the case of traditional binary encoding technique, it takes long time to converge the optimal solutions and brings about complexity of the systems due to encoding and decoding procedures. However, the ROGAs (real-coded genetic algorithms) do not require these procedures, and the k-means clustering algorithm can avoid global searching space. Thus, this paper proposes a new approach by using their advantages. The proposed method constructs the multiple predictors using the optimal differences that can reveal the patterns better and properties concealed in non-stationary time series where the k-means clustering algorithm is used for data classification to each predictor, then selects the best predictor. After selecting the best predictor, the cluster centers of the predictor are tuned finely via RCGKA in secondary tuning procedure. Therefore, performance of the predictor can be more enhanced. Finally, we verifies the prediction performance of the proposed system via simulating typical time series examples.

• Journal of Electrical Engineering and Technology
• /
• 제13권2호
• /
• pp.998-1007
• /
• 2018

The advent of Wireless Sensor Networks (WSN) has led to their use in numerous applications. Sensors are autonomous in nature and are constrained by limited resources. Designing an autonomous topology with criteria for economic and energy conservation is considered a major goal in WSN. The proposed honey-hive clustering consumes minimum energy and resources with minimal transmission delay compared to the existing approaches. The honey-hive approach consists of two phases. The first phase is an Intra-Cluster Min-Max Discrepancy (ICMMD) analysis, which is based on the local honey-hive data gathering technique and the second phase is Inter-Cluster Frequency Matching (ICFM), which is based on the global optimal data aggregation. The proposed data aggregation mechanism increases the optimal connectivity range of the sensor node to a considerable degree for inter-cluster and intra-cluster coverage with an improved optimal energy conservation.

• 한국정보통신학회논문지
• /
• 제15권12호
• /
• pp.2687-2696
• /
• 2011

해양 센서 네트워크에서 클러스터링 문제는 네트워크 수명 및 확장성을 향상시키기 위한 문제 중 하나이다. 본 논문은 모든 노드가 클러스터 헤더를 통하여 싱크 노드로 데이터를 전송하는데 사용되는 전송 에너지를 최소화할 수 있는 클러스터링 문제의 최적 결과를 찾는 알고리즘을 제안한다. 일반적으로 네트워크에서 노드의 수가 증가하면 적절한 클러스터 구성을 위한 계산량은 급격히 늘어나게 된다. 본 논문에서는 적정한 시간 내에 최적의 클러스터 구성을 위한 유전 알고리즘을 제안하며, 효율적인 검색을 위해 유전 알고리즘의 효과적인 이웃해 생성 동작을 제안한다. 제안된 알고리즘은 전송 에너지와 실행시간 관점에서 성능을 평가하며, 평가 결과에서 제안된 알고리즘이 해양 센서 네트워크에서 클러스터를 효과적으로 구성함을 보인다.

• 한국통신학회논문지
• /
• 제38B권6호
• /
• pp.446-453
• /
• 2013

무선 센서네트워크에서 클러스터링 기법은 네트워크 확장성과 네트워크 수명 연장에 효율적이라고 인정받고 있다. 본 논문에서는 클러스터 기반 센서 네트워크에서 multi-hop to one-hop 전송 환경을 고려하여 에너지 효율성에 최적인 클러스터 비율(cluster ratio, CR)을 분석하는데 초점을 둔다. 본 논문에서는 지정한 클러스터 비율을 통한 시스템 홉 수(hop-count) 최소화와 노드 간 패킷수신율(packet reception ratio, PRR) 최대화 사이의 이해득실(trade-off) 관계를 분석하고 이 두 요소를 종합적으로 고려하여 목표함수를 유도한다. 제안한 목표함수를 통하여 얻은 최적의 클러스터 비율은 네트워크에서 패킷 전송에 드는 비용뿐만 아니라 노드 간 재전송 오버헤드를 줄여줌으로써 에너지 효율성을 향상시킨다. 본 논문에서 제안한 기법은 최소 홉 수 클러스터링 방안과 비교되며 시뮬레이션을 통하여 향상된 에너지 효율성을 검증하였다.

• 산업경영시스템학회지
• /
• 제46권4호
• /
• pp.63-73
• /
• 2023

This study explores modern portfolio theory by integrating the Black-Litterman portfolio with time-series clustering, specificially emphasizing K-shape clustering methodology. K-shape clustering enables grouping time-series data effectively, enhancing the ability to plan and manage investments in stock markets when combined with the Black-Litterman portfolio. Based on the patterns of stock markets, the objective is to understand the relationship between past market data and planning future investment strategies through backtesting. Additionally, by examining diverse learning and investment periods, it is identified optimal strategies to boost portfolio returns while efficiently managing associated risks. For comparative analysis, traditional Markowitz portfolio is also assessed in conjunction with clustering techniques utilizing K-Means and K-Means with Dynamic Time Warping. It is suggested that the combination of K-shape and the Black-Litterman model significantly enhances portfolio optimization in the stock market, providing valuable insights for making stable portfolio investment decisions. The achieved sharpe ratio of 0.722 indicates a significantly higher performance when compared to other benchmarks, underlining the effectiveness of the K-shape and Black-Litterman integration in portfolio optimization.

• Journal of information and communication convergence engineering
• /
• 제6권2호
• /
• pp.117-121
• /
• 2008

In ad hoc network, the scarce energy management of the mobile devices has become a critical issue in order to extend the network lifetime. Therefore, the energy consumption is important in the routing design, otherwise cluster schemes are efficient in energy conserving. For the above reasons, an Energy conserving Context aware Clustering algorithm (ECC) is proposed to establish the network clustering structure, and a routing algorithm is introduced to choose the Optimal Energy Routing Protocol (OERP) path in this paper. Because in ad hoc network, the topology, nodes residual energy and energy consuming rate are dynamic changing. The network system should react continuously and rapidly to the changing conditions, and make corresponding action according different conditions. So we use the context aware computing to actualize the cluster head node, the routing path choosing. In this paper, we consider a novel routing protocol using the cluster schemes to find the optimal energy routing path based on a special topology structure of Resilient Ontology Multicasting Routing Protocol (RODMRP). The RODMRP is one of the hierarchical ad hoc network structure which combines the advantage of the tree based and the mesh based network. This scheme divides the nodes in different level found on the node energy condition, and the clustering is established based on the levels. This protocol considered the residual energy of the nodes and the total consuming energy ratio on the routing path to get the energy efficiently routing. The proposed networks scheme could get better improve the awareness for data to achieve and performance on their clustering establishment and messages transmission. Also, by using the context aware computing, according to the condition and the rules defined, the sensor nodes could adjust their behaviors correspondingly to improve the network routing.

• 전기학회논문지
• /
• 제65권4호
• /
• pp.555-560
• /
• 2016

Recently, many countries have placed great attention on energy security and climate changes. Governments are promoting the construction of renewable energy projects with regulatory support in Korea. Despite an increasing penetration of renewable resources, however, the photovoltaic and wind power are underutilized due to the endemic problems such as difficulties of output control and intermittent output. The Energy Storage System (ESS) is proposed as a good solution for solving the problems and has been studied in both the private business and the government. However, because of inefficient aspects, the research has been carried out for improving high costs and a small capacity. In addition, the ESS is currently installed for using only one purpose which is frequency regulation or transmission congestion relief such that has an economic limitation. Therefore, methods which are becoming economically justifiable to increase the penetration of the ESS is required. Thus, this paper presents in terms of operation efficiency to improve economic feasibility of the ESS currently used. mainly, there are two aspects for the operation efficiency. Firstly, it is intended to improve the utilization rate through a process that can utilize the ESS for various purposes. It is necessary to be able to use for other purposes by classifying and clustering for increasing the efficiency of availability. The clustering method is proposed to conduct the grouping the ESS. Especially, it is proposed to utilize ESS for frequency regulation service which is the one of ancillary services in the power system. Through case studies, it is confirmed to secure the necessary resources by clustering small size ESS.