厦门大学 电子工程系 2006级/2010届 本科毕业论文1

厦门夜学 电子工程系 2006级/2010届 本科结业论文

[edit] 蔡利容

• 题 目：我国经济波动的趋向周期分化——基于HP 滤波和状况空间模子的实证剖析
• 摘 要：经济时刻序列中包含不成雅测的时刻趋向成分和周期波动成分，剖析经济波动问题，一个主要的体例就是对经济时刻序列进行趋向周期分化，将时刻趋向序列和 周期波动序列分手出来自力雅察，以获得对经济现象的客雅熟悉。本文考试运用采用HP 滤波和状况空间的趋向周期分化体例，对我国更始开放后的季度GDP 数据进行趋向周期分化，两种分化体例获得的周期成分具有近似的统计特征。基于残差序列的白噪声磨练，对两种体例进行剖析斗劲。对于EViews5.0 软件的运行功效，本文联系现实的经济现象进行了根基剖析。
• 关头词：趋向周期分化, HP滤波, 状况空间模子
• Abstract：The trends and the periodic fluctuations in economic time series are sometimes unobservable variables hidden in the time series. One way to analyze weak economic fluctuations in a strong trend is to decompose the time-trends and the periodic fluctuations and then work with the two time series independently. In this thesis the HP filter and the state space model are used to decompose the time-trend series and the periodic fluctuations hidden in the quarterly GDP of China. Both methods give similar and reliable results. Based on the white noise test of the residuals，both methods are compared.
• Key words：time series trend decomposition, HP filter, state space model.
• 论文

http://www.yangsky.com/researches/teams/2010/EED06/clr/t.pdf

• ppt

http://www.yangsky.com/researches/teams/2010/EED06/clr/ppt.pdf

[edit] 胡铭

• 题 目：数据挖掘及其工业应用
• 摘 要：数据挖掘从理论研究到产物开发只用了短短数年，今朝在国内外已经进入到应用阶段。数据挖掘手艺的应用十分普遍，从政府打点妄想，商业经营，科学研究到 工业妄想撑持等各个规模都可以找到数据挖掘的用武之地。数据挖掘夜体上有两种功能，即展望/验证功能和描述功能。前者指用数据库的若干已知属性展望或验证 其他位置属性值，后者指找到描述数据的可理解模式。近年来，聚类算法作为一种根基的数据挖掘体例被普遍的应用于相似搜索，顾客划分，趋向剖析，金融投资， 地舆信息系统，遥感图像，和信息检索等规模中。本文将着重介绍几种较为有用的聚类算法，并以股票数据和医疗数据对几种聚类算法进行验证，并斗劲其有用性。
• 关头字：数据挖掘，聚类算法，应用，MATLAB
• Abstract：It only took data mining researches a few years to develop from theory to commercial products and services. Currently in China and the rest of the world, the data mining technology has entered into application stage. Data mining technology is widely applied, from management decision, business management, decision support, scientific research to industry and many other fields.There are two functions of data mining in general, that prediction / validation and description of functions. The former refers to a number of known properties of the database to predict or verify the property values of other locations, the latter means finding comprehensible description of the data model. In recent years, clustering algorithms as a basic data mining methods have been widely used in similarity searches, customer classification, trend analysis, financial investment, geographic information systems, remote sensing images, and information retrieval areas.This thesis will focus on several effective clustering algorithms, and apply several clustering algorithms to stock data and medical data to verify and compare their effectiveness.
• Key words：Data Mining， Clustering Algorithms， Matlab
• 论文

http://www.yangsky.com/researches/teams/2010/EED06/hm/t.pdf

[edit] 高秀舒

• 题 目：计较动词轨则在情感建模中的应用初探
• 摘 要：跟着计较动词理论的不竭成长，计较动词已应用于如医学、经济以及人工智能等多个规模。计较动词之所以有如斯快的成长是因为它填补了现代采用静态数据研 究而不是动态描述的欠缺。本文首要探讨计较动词在人工智能情感建模方面的应用，打破传统静态采集数据进行恍惚仿真的体例，采用一种动态的描述体例进行情感 模拟仿真。人工智能的成长一向以来勾留着机械智能的水平上，只懂得从命与执行。本文将介绍了计较动词的根基理论以及首要研究体例，按照动词理论的研究体例 成立情感轨则，并举例说了然在后续仿真工作中的量化体例。最后，本文提出了基于计较动词理论的情感建模的改良体例以及研究目的。本文的内容成立在夜量心理 学资料以及动词理论的基本之上，过程对情感心理学研究的总结归纳得出斗劲合适现实的情感轨则。可是人类激情是一个复杂的过程，是以，本文在数据采集、现实 应用方面还需要做更多的工作，还需不竭地对模子进行改良。
• 关头字：计较动词，情感建模，情感轨则，量化
• Abstract： With the continuous development of computational verb theory, its application involves many fields, such as medical instruments, economics, and artificial intelligence and so on. The reason for the wide applications and deep influences of computational verb theory is its capacities of modeling dynamics. This paper explores the application of computational verb theory to emotional modeling. Psychology has profound influences to our daily life, and yet, many of psychological observations can’t be implemented in the state-of-the-art artificial intelligence. In this paper, a new emotional model based on computational verb theory is built. The concept and modeling structures of computational verb were presented. Besides, it also introduced the process of building an emotional model and gave examples to explain the quantizing process of emotional processes. This paper is built based on big collections of psychological data, which were summarized into emotional rules. Then, these rules are them verbified into computational verb rules.
• Key words：Computational verb, Emotional modeling, Emotion rules, Quantization
• 论文

http://www.yangsky.com/researches/teams/2010/EED06/gxs/t.pdf

• 摘 要：跟着CPU运算能力晋升和自力显卡衬着3D图形的加速，3D游戏已成为游戏开发的主流。游戏公司一般都在Windows下采办商用的或者内部自立研发 的游戏引擎作为游戏开发的框架，这对于游戏开发喜爱者来说价钱过于昂贵。然而免费的游戏引擎文档不齐全，难以搭建起编程平台。本文将介绍若何在Linux 下操作Ogre图形衬着引擎、Newton物理引擎和毗连这两个引擎的OgreNewt库作为工具，开发出两种分歧的3D游戏场景。场景的实现具有精采的 移植性，只要有需要的库文件，也可在Windows下运行。本文首先介绍了编写游戏傍边需要用到的与Ogre、Newton和OgreNewt的相关技 术，然后对这两个游戏场景进行了需求剖析和整体框架的建模，接着编写Makefile剧本把这三个源码的头文件和链接库合成一个“完整”的游戏引擎，最后 讲解了具体的实现细节。具体的实现顶用了多态、担任和模板等手艺降低了模块间的耦合性；使用单件模式和工场模式实现了统一实例化和打点实体，降低了代码的 复杂性。
• 关头字：游戏引擎；3D游戏；游戏场景
• Abstract:3D games have become the main stream of game development with the enhancement of the CPU computing and the acceleration of the 3D graphics rendering by discrete display card. It is too expensive for the lovers of game development that most of game engines today are ready-made to be bought or developed within the independent research. However, the documents of the free game engines are not complete that it is hard to set up the programming platform.This paper introduced how to develop two different types of 3D game scenes racing and shooting games using three engines graphics rendering engine Ogre, physics engine Newton and OgreNewt library that connects these two as the development tool under Linux operating system. Firstly, the Paper introduced the related technologies to Ogre, Newton and OgreNewt used in the game development. Then, requirement analysis and the whole framework modeling of the two game scenes are presented. After that, a Makefile script fusion of these three head files and link libraries made a complete game engine. Finally, the Paper introduced the concrete realization details using polymorphism, inheritance template and other technologies to decreases the degree of coupling between different modules and design patterns such as Singleton, Factory to integrate the instantiations and manage entities that reducing unnecessary code.
• Key words: Game Engine; 3D Game; Game Scene
• 论文

http://www.yangsky.com/researches/teams/2010/EED06/cxq/t.doc

[edit] 江小君

• 题 目：基于计较动词的图像压缩的研究
• 摘 要：本文介绍了计较动词的概念及其在图像压缩规模的应用，并在计较动词图像压缩体例的基本上，引入聚类体例，改良了计较动词图像压缩在尺度动词提取上的不 足，过程该算法措置从原始图像所获得的动词相似度的相关信息，同时应用于在对所获得的尺度动词进行偏移与增益的措置，成立了一个规范的提取尺度动词的方 法。本文所介绍是一种全新的图像压缩体例，虽然压缩比和重构图像质量仍存在不足，但能可以看出计较动词图像压缩是极具潜力的。
• 关头字： 计较动词， 计较动词相似度， 图像压缩， 聚类算法， K-MEANS算法
• Abstract： This thesis introduces the concept of computational verb and its application to image compression . Based on clustering algorithms, in this thesis, the basics of computational verb image compression are presented. By using verb similarity, the original images are represented by the clustering centers in the attribute space constructed from samples of 1D and 2D image regions. In this thesis, different types of sampling windows for constructing observe verbs are studied and compared using Malab simulations. Since the image compression method studied in this thesis is a brand new one, we find that computational verb clustering algorithm is an effective tool though its compression ratio and reconstituted image quality still have a little shortage.
• Key words： computational verb, computational verb similarities, image compression clustering method, K-MEANS
• 论文

http://www.yangsky.com/researches/teams/2010/EED06/jxj/t.doc