全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

基于GDT的对FOIL系统的改进

Keywords: FOIL系统,GDT(generaldistributiontable)方法,规则强度

Full-Text   Cite this paper   Add to My Lib

Abstract:

为了能更有效地处理含有噪音数据的数据集,提出了一个基于GDT(generaldistributiontable)的对FOIL系统的改进方法,该方法利用GDT的思想、规则强度等概念,考虑到数据由噪音引起的不确定性.通过对FOIL系统算法的改进,能够很好地解决FOIL系统对含有噪音训练例集的学习能力,提高FOIL的学习精度.同时,通过例子阐述了该方法的实施过程,分析表明:该算法是一种新的有效地处理含有噪音数据的一阶谓词学习系统.

References

[1]  MITCHELL T M.机器学习[M].北京:机械工业出版社,2003.204-208.MITCHELL TM. Machine Learning[M]. Beijing:China Machine Press,2003. 204-208. (in Chinese)
[2]  QUINLAN J R. Learning logical definitions from relations[J]. Machine Learning, 1990, 5: 239-266.
[3]  QUINLAN J R,CAMERON-JONES R M. Induction of logic programs: FOIL and related systems[J]. New GenerationComputing, 1995, 13:287-312.
[4]  KAMAL Ali. Learning probabilistic relational concept descriptions[D]: [dissertation]. Irvine: University of California,1996.
[5]  ZHONG N, DONG J Z, LIU C, et al. A hybrid model for rule discovery in data[A]. In: ZIARKO W, YAO Y. RoughSets and Current Trends in Computing: Second International Conference[C]. Heidelberg: Springer-Verlag, 2001.436-444.
[6]  ZHONG N, DONG J Z, OHSUGA S. A hybrid approach to rule discovery in databases[J]. Information Sciences, 2000, 126(1-4): 99-127.
[7]  ZHONG N, DONG J Z, OHSUGA S. Data mining based on the generalization distribution table and rough sets[A]. In:WU X. Research and Development in Knowledge Discovery and Data Mining[C]. Berlin: Springer-Verlag, 1998. 360-373.
[8]  ZHONG N, DONG J Z, OHSUGA S. Data mining: A probabilistic rough set approach [A]. In: POLKOWSKI L,SKOWRON A. Studies in Fuzziness and Soft Computing Series[C]. Heidelberg: Physica-Verlag, 1998. 127-146.
[9]  尹旭日,赵志宏,商琳,等.基于GDT的不完整信息系统规则发现[J].计算机研究与发展,2001.38(11):1290-1294.YIN Xu-ri, ZHAO Zhi-hong, SHANG Lin, et al. Rules discovery from incomplete information system by GDT approach[J]. Journal of Computer Research and Development, 2001, 38(11): 1290-1294. (in Chinese)

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133