• Journal of Information Processing Systems
• /
• v.19 no.5
• /
• pp.652-662
• /
• 2023

Identifying highly discriminating genes is a critical step in tumor recognition tasks based on microarray gene expression profile data and machine learning. Gene selection based on tree models has been the subject of several studies. However, these methods are based on a single-tree model, often not robust to ultra-highdimensional microarray datasets, resulting in the loss of useful information and unsatisfactory classification accuracy. Motivated by the limitations of single-tree-based gene selection, in this study, ensemble gene selection methods based on multiple-tree models were studied to improve the classification performance of tumor identification. Specifically, we selected the three most representative tree models: ID3, random forest, and gradient boosting decision tree. Each tree model selects top-n genes from the microarray dataset based on its intrinsic mechanism. Subsequently, three ensemble gene selection methods were investigated, namely multipletree model intersection, multiple-tree module union, and multiple-tree module cross-union, were investigated. Experimental results on five benchmark public microarray gene expression datasets proved that the multiple tree module union is significantly superior to gene selection based on a single tree model and other competitive gene selection methods in classification accuracy.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1997.10a
• /
• pp.170-174
• /
• 1997

This paper studies a method for making more efficient classification rules in the ID3 using the rough set theory. Decision tree technique of the ID3 always uses all the attributes in a table of examples for making a new decision tree, but rough set technique can in advance eleminate dispensable attributes. And the former generates only one type of classification rules, but the latter generates all the possibles types of them. The rules generated by the rough set technique are the simplist from as proved by the rough set theory. Therefore, ID3, applying the rough set technique, can reduct the size of the table of examples, generate the simplist form of the classification rules, and also implement an effectie classification system.

• The KIPS Transactions:PartA
• /
• v.15A no.5
• /
• pp.287-294
• /
• 2008

A tag collision arbitration algorithm for RFID passive tags is one of the important issues for fast tag identification, since reader and tag have a shared wireless channel in RFID system. This paper suggests Hyper-Hybrid Query Tree algorithm to prevent the tag-collisions. The suggested algorithms determine the specified point in time for tag to transfer ID to reader by using value 1 of the upper 3 bit based on Query Tree. Also, because the transferred upper 3 bits of tag is different depending on the time of transfer, it is possible to predict in the suggested Algorithm. In the performance evaluation through simulation, it shows the suggested algorithm has higher performance in the number of queries compared to other Tree-based protocols.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.2
• /
• pp.273-278
• /
• 2010

A few small and medium enterprise administer its organization systematically, but most of them is affected by ability and level of a CEO rather than organization system. In this viewpoint, it can be said the study on ability and level of CEO in small and medium enterprise are so meaningful. Thus, in this paper, the core competency of CEO is obtained from the CEO through questionnaire and it is suggested the evaluation model of the CEO core competency. Also patterns were analyzed by ID3 and fuzzy ID3 from data on expert appraise for CEO core competency and level. The 'if-then' fuzzy rules and decision tree created by results of pattern analysis showed their usefulness for evaluation of CEO core competency in small and medium enterprise.

• 한국HCI학회:학술대회논문집
• /
• 2007.02a
• /
• pp.667-674
• /
• 2007

Facial expressions provide significant clues about one's emotional state; however, it always has been a great challenge for machine to recognize facial expressions effectively and reliably. In this paper, we report a method of feature-based adaptive motion energy analysis for recognizing facial expression. Our method optimizes the information gain heuristics of ID3 tree and introduces new approaches on (1) facial feature representation, (2) facial feature extraction, and (3) facial feature classification. We use minimal reasonable facial features, suggested by the information gain heuristics of ID3 tree, to represent the geometric face model. For the feature extraction, our method proceeds as follows. Features are first detected and then carefully "selected." Feature "selection" is finding the features with high variability for differentiating features with high variability from the ones with low variability, to effectively estimate the feature's motion pattern. For each facial feature, motion analysis is performed adaptively. That is, each facial feature's motion pattern (from the neutral face to the expressed face) is estimated based on its variability. After the feature extraction is done, the facial expression is classified using the ID3 tree (which is built from the 1728 possible facial expressions) and the test images from the JAFFE database. The proposed method excels and overcomes the problems aroused by previous methods. First of all, it is simple but effective. Our method effectively and reliably estimates the expressive facial features by differentiating features with high variability from the ones with low variability. Second, it is fast by avoiding complicated or time-consuming computations. Rather, it exploits few selected expressive features' motion energy values (acquired from intensity-based threshold). Lastly, our method gives reliable recognition rates with overall recognition rate of 77%. The effectiveness of the proposed method will be demonstrated from the experimental results.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.2
• /
• pp.111-116
• /
• 2019

New technologies such as the IoT, Big Data, and Artificial Intelligence, starting from the Web, mobile, and smart device, enable new business models that did not exist before, and various types of Internet companies based on these business models has been emerged. In this research, we examine the factors that influence the success and failure of Internet companies. To do this, we review the recent studies on business model and examine the variables affecting the success of Internet companies in terms of network effect, user interface, cooperation with actors, creating value for users. Using the five derived variables, we will select 14 Internet companies that succeeded and failed in seven commercial business model categories. We derive decision tree by applying inductive learning based on ID3 algorithm to the analysis result and derive rules that affect success and failure based on derived decision tree. With these rules, we want to present the strategic implications for actors to succeed in Internet companies.

• Journal of Information Technology Services
• /
• v.12 no.3
• /
• pp.331-343
• /
• 2013

In RFID systems, the anti-collision algorithm is being improved to recognize Tag's ID within recognition area of the reader quickly and efficiently. This paper focuses on Tag collision. Many studies have been carried out to resolve Tag collision. This paper proposes a new N-ary Query Tree Algorithm to resolve more than Tag collision simultaneously, according to the value of m(2 ~ 6). This algorithm can identify more tags than existing methods by treating a maximum 6 bit collision, regardless of the continuation/non-continuation Tag's ID patterns. So, it extracts maximumly different $2^6$ bit patterns per single prefix in recognition process. The performance of N-ary Query Tree Algorithm is evaluated by theoretical analysis and simulation program.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.5
• /
• pp.700-705
• /
• 2008

In this paper, a model to evaluate the sustainability management for small and middle enterprises was suggested. Also, the if-then rules and its decision tree for pattern analysis which is obtained by fuzzy ID3 from the data of sustainability management were shown. The suggested model can be used for the evaluation tool of competition increasement of enterprises. If the enterprise can recognize that the evaluation rule can be taken advantage of the sustainability management pattern analysis using fuzzy ID3, it is expected that they can use the rule effectively for self evaluation.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.10
• /
• pp.1447-1452
• /
• 2022

Seoul Special City, one of the world's top 10 cities and Metro City, has traditional urban manufacturing industries such as printing, sewing, and mechanical metals. Small business owners in these manufacturing clusters have developed in the form of mutual assistance. Due to the nature of the agglomeration site, each process is handled by an individual company. It is difficult for relatively small business owners to prepare order processing services that provide real-time logistics movement information between processes. This paper collects and analyzes existing logistics data for smooth order and delivery of small business owners in package manufacturing and special printing fields We design an artificial intelligence Fulfillment Service Platform system with CRNN, k-NN, and ID3 Decision Tree Algorithm. Through this study, it is expected that it will greatly contribute to increasing sales and improving capabilities by allowing small business owners in integrated areas to use individual orders and delivery customized services through the Cloud network.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.07b
• /
• pp.625-627
• /
• 2005

결정 트리(Decision Tree)는 주어진 데이터의 경향을 학습하는 데 사용되는 대표적인 방식이다. 이것은 주어진 데이터를 구조화하기 위하여 데이터의 속성과 정보의 엔트로피에 기반을 둔 정보획득량을 이용한다. 본 논문에서는 유비쿼터스 환경에서 사용자 프로파일 정보처럼 시간에 따라 그 경향이 변하는 데이터에 유용하게 적용할 수 있는 시간 가중치 엔트로피를 정의한다. 그리고 ID3 알고리즘을 기반으로 새롭게 제안하는 시간 가중치 엔트로피를 이용하는 향상된 ID3 알고리즘을 쓰고 사용자의 경향을 분석한다. 본 논문에서 제안하는 엔트로피를 이용하는 방식은 데이터들의 시간에 관한 영향을 고려해서 기준방식보다 분석결과가 더욱 유리하다. 두 방식의 비교 테스트 결과를 보면 시간 가중치 엔트로피를 이용하는 알고리즘은 기존의 ID3 알고리즘보다 구성된 트리의 구조가 매우 간단하고 유리하다.