[1]付 敏1,2,戴祖旭,等.压缩编码的上下文树构造算法[J].武汉工程大学学报,2015,37(04):51-55.[doi:10. 3969/j. issn. 1674-2869. 2015. 04. 012]
,,et al.Context tree algorithm based on compression encoding[J].Journal of Wuhan Institute of Technology,2015,37(04):51-55.[doi:10. 3969/j. issn. 1674-2869. 2015. 04. 012]
点击复制
《武汉工程大学学报》[ISSN:1674-2869/CN:42-1779/TQ]
- 卷:
-
37
- 期数:
-
2015年04期
- 页码:
-
51-55
- 栏目:
-
机电与信息工程
- 出版日期:
-
2015-04-30
文章信息/Info
- Title:
-
Context tree algorithm based on compression encoding
- 文章编号:
-
1674-2869(2015).04--0056-03
- 作者:
-
付 敏1; 2; 戴祖旭1; 王道蓬2
-
1武汉工程大学理学院,湖北 武汉 430502;2华中科技大学图像识别与人工智能研究所多谱图像信息处理国防重点实验室,湖北 武汉 430074
- Author(s):
-
FU Min1; 2; DAI Zu-xu1; WANG Dao-peng2
-
1.College of Science, Wuhan Institute of Technology, Wuhan 430502, China; 2.Institute of Pattern Recognition and Artificial Intelligence,Multi-spectral Image Information Processing Key Laboratory of National Defense,Huazhong University of Science and Technology,Wuhan 430074,China
-
- 关键词:
-
上下文树; n元树; 压缩编码
- Keywords:
-
context tree; n_gram tree; compression encoding
- 分类号:
-
TP309.7
- DOI:
-
10. 3969/j. issn. 1674-2869. 2015. 04. 012
- 文献标志码:
-
A
- 摘要:
-
上下文树是构造无算压缩算法的一种重要基础,作为信息处理过程分析随机序列统计特性的常用数据结构,随机序列中的符号来自于某个固定的符号集合. 上下文树一般是一棵n元树,其中n大于1,但是树是一种占用计算机内存较多的数据结构,因此提出了基于压缩编码的上下文树构造算法,根据符号的一阶统计特性对符号做二进制的压缩编码,用二元树代替n(n>2)元树,在相同内存的存储空间下,可以大大增加树的高度. 计算机数值实验表明基于压缩编码的上下树构造对子串做出了更大长度的相关性检测,并且提高了数据分析的精度.
- Abstract:
-
The context tree as a commonly used data structure plays a very important role in analyzing statistical characteristics of random sequence, and the random sequence of symbols generally comes from a fixed symbol set. The general context tree is a n-tree, in which n is more than 1. Because the tree is a kind of computer memory wasting data structure, a context tree construction algorithm based on compress coding was presented utilizing the first-order statistical properties of binary symbols. In the numerical experiment under the same memory storage space condition, the tree’s height has been greatly increased, and the accuracy of data analysis also improved.
参考文献/References:
[1] 洪汉玉,章秀华,程莉.道路病害形态特征的图像分析[J].武汉工程大学学报,2014,36(4):70-76.HONG Han-yu,ZHANG Xiu-hua,CHENG Li. Image analysis method for road disease morphology characteristic[J]. Journal of Wuhan Institute of Technology, 2014,36(4):70-76.(in Chinese)[2] 孙玉昕, 章瑾.利用堆排序优化路径搜索效率的分析[J].武汉工程大学学报,2013,35(10):50-55. SUN Yu-xin, ZHANG Jin.Practical analysis of improving path searching efficiency by heap sort[J].Journal of Wuhan Institute of Technology, 2013,35(10):50-55.(in Chinese)[3] 王学华,刘莉君,马凡杰,等.数控激光加工路径链表快速搜索优化[J].武汉工程大学学报 ,2014,36(10):52-57.WANG Xue-hua,LIU Li-jun,MA Fan-jie et al. Rapid routine searching of numerical control laser processing based on linked list structure[J]. Journal of Wuhan Institute of Technology,2014,36(10):52-57.(in Chinese)[4] 徐超,周一民,沈磊.一种面向隐含主题的上下文树核[J].电子与信息学报,2010, 32(11):2695-2700.XU Chao,ZHOU Yi-min, SHEN Lei.A context tree kernel based on latent semantic topic[J]. Journal of Electronics & Information Technology,2010,32(11):2695-2606.(in Chinese)[5] RISSANEN J. A universal data compression system[J]. IEEE Transactions on information theory, 1983, 29(5): 656-664.[6] 陈亮,孟庆愿, 董彦磊,等.CTW无损压缩算法在管道无损检测中的应用[J].实验技术与管理, 2012,29(6):42-47.CHEN Liang,MENG Qingyuan,DONG Yanlei,et al.Using CTW lossless compression algorithm in pipelines nondestructive testing[J]. Experimental Technology and Management,2012,29(6):42-47.(in Chinese)[7] DUMONT Thierry.Context tree estimation in variable length hidden[J]. IEEE Transactions on information theory, 2014, 60(6): 3196-3208.[8] GWADERA R, GIONIS A, MANNILA H. Optimal segmentation using tree models[J]. Knowledge and Information Systems, 2008, 15(3): 259-283.[9] Martins D A, Neves A J R, Pinho A J. Variable Order Finite-Context Models in DNA Sequence Coding[C]//Pattern Recognition and Image Analysis. Springer Berlin Heidelberg, 2009: 457-464.[10] B?譈HLMANN P. Model selection for variable length Markov chains and tuning the context algorithm[J]. Annals of the Institute of Statistical Mathematics, 2000, 52(2): 287-315.[11] C?魪NAC P, CHAUVIN B, PACCAUT F, et al. Context trees, variable length Markov chains and dynamical sources[C]//Séminaire de Probabilités XLIV. Springer Berlin Heidelberg, 2012: 1-39.
备注/Memo
- 备注/Memo:
-
收稿日期:2015-01-12基金项目:湖北省自然科学基金重点项目(2010CDA009),湖北省自然科学基金一般项目(2009CDB367),国家自然科学基金面上项目(61175013),武汉工程大学校级教研项目(X2013021).作者简介:付敏(1979-),女,湖北襄阳人,讲师,博士研究生.研究方向:信息处理袁计算机视觉.
更新日期/Last Update:
2015-05-25