本文選題：云數(shù)據(jù) + 海量數(shù)據(jù)��； 參考：《計(jì)算機(jī)仿真》2017年04期

【摘要】：云平臺(tái)上海量數(shù)據(jù)中用戶信息的提取,可更好地提升云提取的服務(wù)質(zhì)量。對信息的準(zhǔn)確提取,需要給出數(shù)據(jù)特征淘汰特性和過濾內(nèi)在聯(lián)系性,對數(shù)據(jù)特征進(jìn)行匹配來完成。傳統(tǒng)方法通過統(tǒng)計(jì)樣本數(shù)據(jù)的頻率表,提取每個(gè)數(shù)據(jù)特征的不一致性,但無法互相匹配,導(dǎo)致提取精度低。提出基于改進(jìn)K近鄰的云平臺(tái)上海量數(shù)據(jù)中提取用戶信息數(shù)學(xué)模型。以原始的云數(shù)據(jù)輸人空間的特征為提取因子,對各個(gè)條件數(shù)據(jù)屬性依據(jù)相同的權(quán)重提取特征樣本間的距離,得到不同條件屬性下相應(yīng)特征參數(shù)的聯(lián)合熵,給出數(shù)據(jù)特征淘汰特性和過濾的內(nèi)在聯(lián)系性,采用分?jǐn)?shù)階Fourier變換進(jìn)行數(shù)據(jù)特征的匹配,構(gòu)建了K-L數(shù)據(jù)特征分類器,以上述分類器為依據(jù)組建云平臺(tái)上海量數(shù)據(jù)中提取用戶信息數(shù)學(xué)模型。實(shí)驗(yàn)結(jié)果表明,所提模型提取精確度較高。
[Abstract]:The extraction of user information from cloud platform Shanghai quantity data can improve the service quality of cloud extraction.To extract the information accurately, we need to give the characteristic of data feature elimination and the inherent relation of filtering, and match the data feature to complete.The traditional method extracts the inconsistency of each data feature through the frequency table of statistical sample data, but it can not match each other, which leads to the low precision of extraction.A mathematical model of extracting user information from cloud platform Shanghai quantity data based on improved K nearest neighbor is proposed.With the feature of the original cloud data input into human space as the extraction factor, the joint entropy of the corresponding feature parameters under different condition attributes is obtained by extracting the distance between the feature samples for each conditional data attribute according to the same weight.The intrinsic relation between feature elimination and filtering is given. A K-L data feature classifier is constructed by using fractional order Fourier transform to match data features.Based on the above classifier, a mathematical model of extracting user information from cloud platform Shanghai quantity data is established.The experimental results show that the proposed model has high accuracy.
【作者單位】： 貴州財(cái)經(jīng)大學(xué)數(shù)學(xué)與統(tǒng)計(jì)學(xué)院;
【分類號(hào)】：O141.4;TP393.09
，

本文編號(hào)：1761289