隨著信息技術(shù)的不斷進(jìn)步,日常生活和工業(yè)環(huán)境中的數(shù)據(jù)持續(xù)增長(zhǎng),如何理解這些數(shù)據(jù)的含義從而幫助用戶做出決策成為一項(xiàng)嚴(yán)峻而具有挑戰(zhàn)性的難題。數(shù)值型數(shù)據(jù)在表述不確定信息時(shí),往往無(wú)法達(dá)到對(duì)數(shù)據(jù)的完整性和準(zhǔn)確性的要求,信息粒為解決這類不確定問題提供了更有效的解決方案。信息粒是基于數(shù)據(jù)的特征性和近似性精心設(shè)計(jì)并抽象化的數(shù)據(jù)集合,它可以完整并準(zhǔn)確地表達(dá)數(shù)據(jù)的含義。通過信息粒化,復(fù)雜問題被分解為一系列易于處理的子問題,從而降低問題總體成本。在粒計(jì)算中,如何構(gòu)建信息粒以及如何將信息粒應(yīng)用于生產(chǎn)建模中成為當(dāng)前亟待解決的問題之一。本文旨在通過對(duì)信息粒構(gòu)建方法的研究,建立一種通用的信息粒構(gòu)建模型,并在此基礎(chǔ)上,進(jìn)一步研究信息粒的優(yōu)化以及應(yīng)用。本文從不同方面對(duì)信息粒進(jìn)行概念和算法上的研究,例如高階信息粒,尤其是粒區(qū)間信息粒的構(gòu)建（即信息粒的分層結(jié)構(gòu)）、信息粒的優(yōu)化（粒度數(shù)據(jù)聚合）以及信息粒在模糊規(guī)則模型中的應(yīng)用等。針對(duì)Ⅰ型信息粒描述數(shù)據(jù)時(shí)存在的局限性（比如無(wú)法滿足實(shí)際問題中對(duì)描述數(shù)據(jù)的完整性和準(zhǔn)確性的要求）,本文提出一種應(yīng)用合理粒度準(zhǔn)則構(gòu)建高階信息粒（即粒區(qū)間信息粒）的方法。以往的研究對(duì)應(yīng)用合理粒度準(zhǔn)則構(gòu)建I型信... 

【文章來源】：西安電子科技大學(xué)陜西省 211工程院校 教育部直屬院校

【文章頁(yè)數(shù)】：143 頁(yè)

【學(xué)位級(jí)別】：博士

【文章目錄】：
ABSTRACT
摘要
List of Symbols
List of Abbrevations
Chapter 1 Introduction
    1.1 Backgroud
    1.2 Granular Computing:A Literature Review
    1.3 Contributions
        1.3.1 Organization
        1.3.2 Significiance
Chapter 2 Information Granularity:Basic Concepts
    2.1 Formalisms of Information Granules
        2.1.1 Sets and Intervals
        2.1.2 Fuzzy Sets
        2.1.3 Rough Sets
        2.1.4 Shadowed Sets
        2.1.5 Other Formalisms
    2.2 Information Granules of Higher Type
    2.3 Design Methods of Information Granules
        2.3.1 The Principle of Justifiable Granularity
        2.3.2 Clustering Algorithms
    2.4 Conclusions
Chapter 3 Design of Information Granules of Higher Type and Their Applications to System Modeling
    3.1 Granular Computing:Representing and Describing Data with Information Granules
    3.2 The Principle of Justifiable Granularity:Conceptual Developments and Underlying Generic Algorithm
        3.2.1 Coverage
        3.2.2 Specificity
        3.2.3 Performance Index
    3.3 Development of a Granular Interval
    3.4 Application Studies
    3.5 Granular Characterization of Numeric Models
    3.6 Construction of Multidimentsional Information Granules
    3.7 Conclusions
Chapter 4 Granular Data Aggregation:An Adaptive Principle of Justifiable Granularity Approach
    4.1 A Framework of Data Aggregation
    4.2 Aggregation Mechanisms— A Focused Review
    4.3 Adaptive Principle of Justifiable Granularity
        4.3.1 Formation of A Numeric Representative
        4.3.2 Formation of Information Granules Around Numeric Representative.
        4.3.3 Optimization Process
    4.4 Experimental Studies
        4.4.1 Prediction of Time Series
        4.4.2 Prediction in Spatially Distributed Data
    4.5 Conclusions
Chapter 5 A Two-Phase Design of Fuzzy Rule-Based Model and Its Applications
    5.1 Fuzzy Rule-Based Architecture
    5.2 Prerequisites
        5.2.1 Formation of Fuzzy Sets of Condition and Conclusion
        5.2.2 The Principle of Justifiable Granularity in Developing Information Granules
    5.3 Characterization of the Quality of the Rules
    5.4 Processing in Granular Rule-Based Model
        5.4.1 Formation of Granular Rule-Based Model
        5.4.2 Evaluation of Performance of Granular Rule-Based Model
    5.5 Experimetal Studies
        5.5.1 Synthetic Experiments
        5.5.2 Analysis of Fuzzy Rule-Based Model
    5.6 Conclusions
Chapter 6 Conclusions and Future Research
    6.1 Conclusions
    6.2 Future Research
References
Aknowledgement
Biography



本文編號(hào)：3265856