Font Size: a A A

Pattern Recognition Method&Its Application To Classification Of Water Quality Etc

Posted on:2005-06-25Degree:MasterType:Thesis
Country:ChinaCandidate:W LuFull Text:PDF
GTID:2121360125466506Subject:Applied Chemistry
Abstract/Summary:PDF Full Text Request
Chemistry and its relative subjects had many variables and parameters. Many metrical datum should be dealed with through multianaylsis and classified. Computer Pattern Recognition was introduced. Pattern Recognition had strong capability of learning and fitness which could overcome contrived factor and classify samples correctly and quickly.Statistical Pattern Recognition of Pattern Recognition was expatiated. The method of classification and cluster was emphasized. To realize the classification and cluster of samples, Pretreatment should be taken at first. Then feature selection and extraction was carried out. Next corresponding discriminant function was established.Classification decision was maken finally. By editing Qbasic program,classification was realized. But cluster of samples was through SPSS software. Two-class surface water was classified through Qbaisc. Vinegar tea and circulating water quality.Linear learning method and K-Nearest Neighbour Method was used to classify surface water of Nantong and Sheyang. The recognition ratio and prediction ratio reached 100%.Hierarchical cluster method was used to cluster green tea and red tea .The result proved well.Three components were extracted through principal component analysis. Then hierarchical cluster method and K-means method was used to cluster vinegar of Zhenjiang & Shanxi. The result was the same entirely. Explanation was set forth to samples of wrong classified.The datum of circulating supply water quality was centralized. Principal component matrix was rotated through factor analysis. Hierarchical cluster of multiple class was realized through factor after rotation. Key factor of influencing cluster was searched. The recipe was recommended and worked well.
Keywords/Search Tags:pattern recognition, classification, cluster, principal component analysis, factor analysis
PDF Full Text Request
Related items