首页>
外国专利>
Mixed data fingerprinting by principal component analysis
Mixed data fingerprinting by principal component analysis
展开▼
机译:通过主成分分析的混合数据指纹
展开▼
页面导航
摘要
著录项
相似文献
摘要
Principal component analysis is applied to the data set to fingerprint the data set or to compare the data set to a "wild file" that may be made up of data found in the data set. Principal component analysis makes it possible to reduce the data set of the data used for comparison into sparingly compressed signatures. Data sets with different patterns between the variables will have different pattern principal components. The principal components of the variables (or a related subset thereof) in the wild file can be calculated and statistically compared to the principal components of the same variable in the data provider's reference file to provide a score. This constitutes a unique and compressed signature of the file that can be used for identification and comparison with similarly defined patterns from other files.
展开▼