朴素贝叶斯在笔记本代工行业中的应用

发布时间：2018-02-03 01:35

本文关键词： 模式识别笔记本电脑生产产能朴素贝叶斯算法生产产能因素　出处：《上海交通大学》2012年硕士论文　论文类型：学位论文

【摘要】：自上世纪90年代起，笔记本电脑以其轻便、高效、灵活等特点逐步成为人们生活中的必需品。随着用户对市场上笔记本的更新速度、稳定性和价格等多方面需求的不断提高，笔记本代工制造行业的竞争也变得异常激烈。技术平台的更新使得笔记本研发更新换代的速度以“月”为单位，，高速、高效、利润最大化的理念已经成为笔记本电脑代工制造行业间竞争生存的不二法宝。快速反应，高产能，过硬质量，强大物流及成本优势是笔记本代工制造行业发展的必要条件。如何能按照笔记本品牌商的要求，及时保证大批量物美价廉的笔记本按计划上市，是笔记本代工制造行业目前最重要的课题。结合这些来自笔记本代工生产企业的实际需求，我们针对某笔记本电脑制造企业的生产状况进行了调查。根据在笔记本代工企业多年相关的实际工作经验和企业数据库中提取出与笔记本电脑生产运作流程有关的数据信息，我们初步总结出三个主要影响制造企业日常生产运作的因素包括：人为因素、生产计划因素和设备因素。基于此假定，本文通过来自笔记本制造企业的大量实际数据应用朴素贝叶斯（Naive Bayesian）分类算法进行验证分析。本文所述结果可帮助笔记本代工制造企业在做出生产资源配置决策，辅助企业制定更为合理、准确和科学的生产运作制度，以达到企业利润的最大化，并赢得更多客户和订单的目的。本项目所涉及的具体工作包含： (1)梳理笔记本代工企业生产制造流程，并对影响企业日常生产因素进行设。 (2)学习并研究模式识别及模式识别中的朴素贝叶斯算法。 (3)基于朴素贝叶斯算法理论设计实现了可应用于某笔记本代工企业的软件，并对其中训练样本加载及执行流程做了详细的阐述。且对算法有效性做出评估，从而分析出真正影响笔记本生产的因素。 (4)根据实验结果，对朴素贝叶斯算法应用于笔记本代工企业日常生产的可行性进行了论证，并对笔记本生产产能的预测做出总结，验证算法的有效性。本项目的实验结果基本印证了文章提出的假定，也同时证明了该算法在对笔记本代工制造企业产能分析上的有效性。
[Abstract]:Since -10s, notebook computers with its portable, efficient, flexible and other characteristics have gradually become a necessity in people's lives. With the increasing demand for stability and price, the competition in the notebook contract manufacturing industry has become extremely fierce. The update of the technology platform makes the speed of notebook R & D update to the "month" as the unit, high speed. The concept of high efficiency and profit maximization has become the key to the survival of the notebook computer contract manufacturing industry. Rapid response, high capacity, excellent quality. Strong logistics and cost advantages are the necessary conditions for the development of notebook contract manufacturing industry. How to ensure that large quantities of high-quality and inexpensive notebooks can be listed as planned according to the requirements of notebook brands. It is the most important subject of notebook contract manufacturing industry at present. Combine these from the actual needs of notebook contract manufacturing enterprises. We investigated the production status of a notebook computer manufacturing enterprise. Based on the actual work experience and enterprise database of notebook contract manufacturing enterprise for many years, we extracted from the notebook computer production operation process. Data information. We preliminarily summarize three main factors that affect the daily production operation of manufacturing enterprises: human factors, production planning factors and equipment factors. Based on this assumption. This paper applies naive Bayesian Bayesian (Bayesian) to a large number of practical data from notebook manufacturing enterprises. The results of this paper can help the notebook manufacturing enterprises to make the decision of allocation of production resources. Assist enterprises to establish more reasonable, accurate and scientific production operation system to maximize profits and win more customers and orders. The specific work involved in this project includes: 1) carding the manufacturing process of Notebook manufacturer, and setting up the daily production factors. Learning and studying naive Bayes algorithm in pattern recognition and pattern recognition. Based on the theory of naive Bayes algorithm, the software which can be used in a notebook contract manufacturing enterprise is designed and implemented. The training sample loading and execution flow are described in detail, and the effectiveness of the algorithm is evaluated, and the factors that affect the notebook production are analyzed. 4) based on the experimental results, the feasibility of applying naive Bayes algorithm to the daily production of notebook contract manufacturing enterprises is demonstrated, and the prediction of notebook production capacity is summarized to verify the validity of the algorithm. The experimental results of this project basically confirm the hypothesis put forward in this paper, and at the same time prove the validity of the algorithm in the capacity analysis of notebook contract manufacturing enterprises.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP368.32

【参考文献】