• Title/Summary/Keyword: ID3 알고리즘

Search Result 67, Processing Time 0.022 seconds

Analysis of Stability Condition and Wideband Characteristics of 3D Isotropic Dispersion(ID)-FDTD Algorithm (3차원 ID-FDTD 알고리즘의 Stability Condition과 광대역 특성 분석)

  • Kim, Woo-Tae;Koh, Il-Suek;Yook, Jong-Gwan
    • The Journal of Korean Institute of Electromagnetic Engineering and Science
    • /
    • v.22 no.4
    • /
    • pp.407-415
    • /
    • 2011
  • The stability condition and wideband characteristics of 3D ID-FDTD algorithm which has low dispersion error with isotropic dispersion are presented in this paper. 3D ID-FDTD method was proposed to improve the defect of the Yee FDTD such as the anisotropy and large dispersion error. The published paper calculated the stability condition of 3D ID-FDTD algorithm by using numerical method, however, it is thought that the examples were not sufficient to verify the stability condition. Thus, in this paper, various simulations are included in order to hold reliability under the conditions that the plane wave propagation is assumed with a single frequency and a wideband frequency. Also, the 3D ID-FDTD algorithm is compared to those that have the similar FDTD algorithm with ID-FDTD such as Forgy's method and non-standard FDTD method in a wideband. Finally, the radar cross section(RCS) for the large sphere with high dielectric constant is calculated.

ID3 Algorithm Improved with Time-weighted Entropy (시간 가중치 엔트로피를 이용한 결정 트리 생성 알고리즘)

  • Dong, Liq-Uan;Lee, Jee-Hyong
    • 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.

  • PDF

ID-Based Cryptosystem and Digital Signature Scheme Using Discrete Logarithm Complexity (이산대수 문제를 이용한 ID 기본 암호시스템과 디지틀 서명방식에 관한 연구)

  • 염흥열
    • Journal of the Korea Institute of Information Security & Cryptology
    • /
    • v.1 no.1
    • /
    • pp.3-15
    • /
    • 1991
  • in 1984 Shamir proposed the concept of ID(identity)-based cryptosystem and digital signature scheme wich does not require any public file for storing the user' s public keys. In this paper, the concept of the ID-based cryptosystem based on discrete logarithm problem using the conven-tional public-key cryptosystems is described, and the implementation method of the ID-based cryptosystem is also presented. The new digital signaturw schme based on the ID-based crypto-system is proposed and possible atacks are considered and analyzed for the security of digital signature scheme. The proposed ID-based schemes are particularrly useful for smart card and personal identification card application.

Research on Success and Failure of Mobile operating system using inductive learning based on ID3 algorithm (ID3 알고리즘 기반의 귀납적 추론을 활용한 모바일 OS의 성공과 실패에 대한 연구)

  • Jin, Dong-Su
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.19 no.2
    • /
    • pp.258-264
    • /
    • 2015
  • This research suggests critical factors affecting success and failure of several commercial mobile operating systems from Palm OS appearing in to main mobile OSs appearing in 2013. For this, we analyses several mobile operating cases, elicit factors affecting success and failure of mobile OS, and conduct ID3 based inductive learning analyses based on elicited factors and values in case dataset. Through this, we draw rules in success and failure of mobile OS and suggest strategic implications for the commercial success of mobile OS in perspective of innovation and globalization.

Data Mining using ID3 (ID3를 활용한 데이터 마이닝)

  • 석현태
    • Proceedings of the Korea Institute of Convergence Signal Processing
    • /
    • 2003.06a
    • /
    • pp.38-41
    • /
    • 2003
  • There are many kinds of algorithms used for the purpose of data mining. But without the understanding the underlying principles in the algorithm, the result of the data mining cannot be interpreted correctly. In this paper, the principle of ID3 algorithm is explained for that purpose. In addition, the way how to generate good training examples from the relational database is treated, as well as how to convert continuous values into discrete values is considered to use the algorithm for the data mining of real world database.

  • PDF

Enhanced VLC-TDoA Algorithm for Indoor Positioning Without LED-ID (LED-ID 없이 실내 위치 추정이 가능한 개선된 VLC-TDoA 알고리즘)

  • Do, Trong-Hop;Hwang, Junho;Yoo, Myungsik
    • The Journal of Korean Institute of Communications and Information Sciences
    • /
    • v.38B no.8
    • /
    • pp.672-678
    • /
    • 2013
  • In the recent year, along with the rapid development of LED technology, many applications using LEDs have been researched and indoor positioning is one of them. In particular, previous indoor positioning systems based on visible light communication combine triangulation manner such as AoA, ToA TDoA. But most of them needs transmitting unique ID of each LED panel. In this paper, we propose an non LED-ID based indoor positioning system in which the visible light radiated from LEDs is used to locate the position of receiver. Compared to current indoor positioning systems using LED light, our system has the advantages of simple implementation, low cost and high accuracy. Through the computer simulation, our system can achieve a high estimation accuracy of 3.6cm in average in the $5{\times}5{\times}3m^3$ room.

The development of RFID multi-codes converter based on ID profiles (ID 프로파일을 이용한 RFID 멀티 코드 변환기 연구)

  • Lee, Chang-Yeol;Mo, Hee-Sook
    • Journal of the Institute of Electronics Engineers of Korea CI
    • /
    • v.46 no.2
    • /
    • pp.124-133
    • /
    • 2009
  • There are many different ID representation forms depending on the media or applications. In case of RFID tag, ID representation form must be followed by the rule of ISO/IEC 15962. In this study, we developed the efficient ID conversion algorithm between ID representation form on RFID tag and Internet. The main idea is on the use of XML based ID profiles and three step logical IDs forms. The algorithm was tested by the typical three kinds of real IDs such as EPC, ISO/IEC 15459 KKR Code, and mCode which are the typical meta-IDs can be defined in ISO/IEC 18000-6C tag.

Designing Neural Network Using Genetic Algorithm (유전자 알고리즘을 이용한 신경망 설계)

  • Park, Jeong-Sun
    • The Transactions of the Korea Information Processing Society
    • /
    • v.4 no.9
    • /
    • pp.2309-2314
    • /
    • 1997
  • The study introduces a neural network to predict the bankruptcy of insurance companies. As a method to optimize the network, a genetic algorithm suggests optimal structure and network parameters. The neural network designed by genetic algorithm is compared with discriminant analysis, logistic regression, ID3, and CART. The robust neural network model shows the best performance among those models compared.

  • PDF

Efficient Digital Signal Processing of DTV TxID Based On Decimation Algorithm (Decimation 알고리즘을 이용한 DTV TxID수신 신호의 효율적인 신호처리 기법에 관한 연구)

  • Son, Ju-Hee;Lee, Yong-Tae;Park, Sung-Ik;Kim, Heung-Mook;Seo, Jae-Hyun;Lee, Jae-Young;Cha, Jae-Sang
    • Journal of Broadcast Engineering
    • /
    • v.12 no.3
    • /
    • pp.266-277
    • /
    • 2007
  • Recently, Transmitter Identification(TxID) technology has been issued as a technology of ATSC DTV. ATSC DTV networks are comprised of a plurality of transmitters, broadcasting the same signal sing one frequency network(SFN) connected to EDOCR. In this single frequency network, TxID technology has been recognized as a technology in the ATSC DTV system since it enables the broadcast authorities and classify multiple transmitters. However, conventional TxID uses extremely long spreading sequence to identifying transmitters, so it increases H/W complexity and registers. Thus, to solve those hardware problems, we propose an efficient signal processing technology using decimation algorithm. Furthermore, we certified the availability of the proposed algorithm via various simulations.