共查询到20条相似文献,搜索用时 46 毫秒
1.
本文研究了一种数据驱动下的半导体生产线调度框架,该框架基于调度优化数据样本,应用机器学习算法,获得动态调度模型,通过该模型,对于半导体生产线,能够根据其当前的生产状态,实时地定出近似最优的调度策略.在此基础上,利用特征选择和分类算法,提出一种生成动态调度模型的方法,并且具体实现出一种混合式特征选择和分类算法的调度模型:先采用过滤式特征选择方法对生产属性进行初步筛选,然后再采用封装式特征选择和分类方法生成模型以提高模型生成的效率.最后,在某实际半导体生产线上,对在所提出的框架上采用6种不同算法实现的动态调度模型进行测试,并对算法性能数据和生产线性能据进行对比和分析.结果表明,数据驱动下的动态调度方法优于单一的调度规则,同时也能满足生产线调度实时性要求.在数据样本较多的情况下,建议采用本文所提出的方法. 相似文献
2.
数据驱动的集成化工过程系统 总被引:4,自引:4,他引:4
提出数据驱动的集成化工过程系统(ICPS)的功能和总体结构.开发了适合化工过程模拟的数据库管理子系统,开发了面向数据库库化工单元模块和过程模块,按面向对象的策略在微机上实现了集成化工过程系统. 相似文献
3.
供应链管理企业和行业的数字化是当前重要的发展方向,但传统的供应链管理案例研究方法主要关注案例创新和案例结果,研究偏宏观和定性。文章基于数据和模型双重驱动的视角,研究新的供应链管理案例方法,构建包括网络结构、业务流程、功能定位、绩效评估和战略评估五个部分的总体框架,并提出对应的分析方法,同时,提供盒马鲜生案例研究实例。该方法可提高供应链管理案例研究的质量。 相似文献
4.
基于模型驱动的数据映射技术研究 总被引:1,自引:0,他引:1
给出了数据词典的逻辑结构,以数据词典的方式在元模型层上对数据模型进行统一描述,在数据词典的基础上分析了异构数据模型之间的数据映射关系。采用EBNF范式对描述映射关系的语法进行形式化定义和描述,给出映射关系的相关语义说明,为映射关系可视化建模提供图形符号元素的表示方法,使异构数据间的集成以模型驱动的方式实现。 相似文献
5.
6.
本文基于数据驱动机制,提出了一种描述数据流程图DFD的中间语言并实现了对它的编译,提出了关系矩阵,行列向量表示法等,采用双缓冲,递归下降子程序等方法实现了各种必需的编译分析及目标代码的生成,是一种有效的辅助MIS开发的工具。 相似文献
7.
8.
9.
11.
Raghu Ramakrishnan 《Constraints》1997,2(1):93-94
Recently, interest in constraints and databases has increased significantly as the databases field has started to consider broader application domains. In this short position statement, I identify some of the most promising direcrtions, in my opinion, for the use of constraints in the context of databaes management systems. 相似文献
12.
13.
针对目前多分类代价敏感分类问题在转换成二分类代价敏感分类问题存在的代价合并问题, 研究并构造出了可直接应用于多分类问题的代价敏感AdaBoost算法.算法具有与连续AdaBoost算法 类似的流程和误差估计. 当代价完全相等时, 该算法就变成了一种新的多分类的连续AdaBoost算法, 算法能够确保训练错误率随着训练的分类器的个数增加而降低, 但不直接要求各个分类器相互独立条件, 或者说独立性条件可以通过算法规则来保证, 但现有多分类连续AdaBoost算法的推导必须要求各个分类器相互独立. 实验数据表明, 算法可以真正实现分类结果偏向错分代价较小的类, 特别当每一类被错分成其他类的代价不平衡但平均代价相等时, 目前已有的多分类代价敏感学习算法会失效, 但新方法仍然能 实现最小的错分代价. 研究方法为进一步研究集成学习算法提供了一种新的思路, 得到了一种易操作并近似满足分类错误率最小的多标签分类问题的AdaBoost算法. 相似文献
14.
15.
基于放大因子和延伸方向研究流形学习算法 总被引:16,自引:0,他引:16
流形学习是一种新的非监督学习方法,可以有效地发现高维非线性数据集的内在维数和进行维数约简,近年来越来越受到机器学习和认知科学领域研究者的重视.虽然目前已经出现了很多有效的流形学习算法,如等度规映射(ISOMAP)、局部线性嵌套(Locally Linear Embedding,LLE)等,然而,对观测空间的高维数据与降维后的低维数据之间的定量关系,尚难以直观地进行分析.这一方面不利于对数据内在规律的深入探察,一方面也不利于对不同流形学习算法的降维效果进行直观比较.文中提出了一种方法,可以从放大因子和延伸方向这两个方面显示出观测空间的高维数据与降维后的低维数据之间的联系;比较了两种著名的流形学习算法(ISOMAP和LLE)的性能,得出了一些有意义的结论;提出了相应的算法从而实现了以上理论.对几组数据的实验表明了研究的有效性和意义. 相似文献
16.
In this paper, we consider discrete event logistics systems (DELS). DELS are networks of resources through which material flow. They have been the subject of a large body of analytic research. A huge variety of specific models exists that generally require application by model and/or solution experts to answer narrowly-defined logistics questions about inventory, sourcing, scheduling, routing, etc. It has, however, proven difficult to integrate these models in any comprehensive way into information systems for logistics because of the lack of conceptual alignment between the models produced by researchers and the information systems deployed in practice with which they should be integrated. In this paper, we systematically identify several challenges for DELS research. We analyse the root of the difficulties for applying academic research results in DELS, and indicate some potential future research directions. 相似文献
17.
18.
Mapping problems are typical research topics related to natural language learning, and they include not only classification mappings but also nonclassification mappings, such as verbs and their past tenses. Connectionist computational models are one of the most popular approaches for simulating those mapping problems; however, their lack of explanatory ability has prevented them from being further used to understand the language learning process. Therefore, the work of extracting rational rules from a connectionist model is as important as simulating the mapping behaviors. Unfortunately, there is no available technique that can be directly applied in those computational models to simulate nonclassification problems. In this paper, an autoencoder‐based connectionist computational model is proposed to derive a rule extraction method that can construct “If‐Then” rational relations with high fidelity for nonclassification mapping problems. To demonstrate its generalizability, this computational model is extended to a modified version to address a multi‐label classification mapping problem related to cognitive style prediction. Experiments prove this computational model's simulation ability and its explanatory ability on nonclassification problems by comparing its fidelity performances with those of the classical connectionist computational model (multilayer perceptron artificial neural network), and its similar ability on a multi‐label classification problem (Felder‐Silverman learning style classification) by comparing its prediction accuracy with those of other rule induction techniques. 相似文献
19.
An optimal placement problem for special-form objects on a multiply connected domain is considered. The objects must be disjoint
and belong to a convex region that contains forbidden areas. The performance criterion is described by a function of maximum
of differentiable functions. The method of feasible directions and the method of directed transition were adapted to solve
this problem.
__________
Translated from Kibernetika i Sistemnyi Analiz, No. 3, pp. 128–137, May–June 2007. 相似文献
20.
针对用于人脸识别的主分量分析(PCA)计算量大、识别效果不佳和分类时间长,线性判别分析(LDA)存在小样本问题,比较研究几种基于PCA和LDA的人脸识别方法,这些方法包括PCA+ LDA、2DPCA+ DLDA和2DPCA+2DLDA.在理论和实验上比较研究这些方法,且在ORL和Yale人脸库的实验数据表明,结合后的方法比结合前的方法分类效果好,分类时间短,且在这几种方法中,2DPCA+2DLDA的效果最佳. 相似文献