.png){kind=icon}
返回
Mining in chemometrics
.png){kind=icon}
PRE
AI
DOI：10.1016/j.aca.2008.02.025.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Some of the increasingly spread data mining methods in chemometrics like exploratory data analysis, artificial neural networks, pattern recognition, and digital image processing with their highs and lows along with some of their representative applications are discussed. The development of more complex analytical instruments and the need to cope with larger experimental data sets have demanded for new approaches in data analysis, which have led to advanced methods in experimental design and data processing. Hypothesis-driven methods typified by inferential statistics have been gradually complemented or even replaced by data-driven model-free methods that seek for structure in data without reference to the experimental protocol or prior hypotheses. The emphasis is put on the ability of data mining methods to solve multivariate-multiresponse problems on the basis of experimental data and minimal statistical assumptions only, in contrast to classical methods, which require predefined priors to be tested against some null-hypothesis. (C) 2008 Elsevier B.V. All rights reserved.
Keyword:
chemometrics
data mining
exploratory analysis
pattern recognition
artificial neural networks
inferential statistics
hypothesis-driven methods
data-driven methods
期刊
IF:
6
论文数:
3.3W
被引数:
6.1W
机构
引用论文
.png){kind=icon}
Additions and corrections: Potential energy models for nonbonding interactions in solid chlorine
PREAI
分享收藏分享收藏
分享收藏
查看更多内容