本文選題：微博平臺(tái) + 特征信息�。� 參考：《計(jì)算機(jī)仿真》2017年05期

【摘要】：對(duì)微博平臺(tái)特征信息優(yōu)化檢測(cè)的研究,可對(duì)海量微博信息中所需信息進(jìn)行高效檢索。對(duì)特征信息優(yōu)化檢測(cè)的過(guò)程,需要對(duì)信息重排,并進(jìn)行主成分特征提取,完成對(duì)特征信息的檢測(cè)。傳統(tǒng)方法結(jié)合相關(guān)性檢驗(yàn),對(duì)樣本信息流進(jìn)行處理和分析,但忽略了對(duì)信息的主成分進(jìn)行特征提取,導(dǎo)致檢測(cè)精度偏低。提出基于螢火蟲(chóng)優(yōu)化神經(jīng)網(wǎng)絡(luò)的微博平臺(tái)特征信息優(yōu)化檢測(cè)仿真。采用自回歸移動(dòng)平均模型對(duì)統(tǒng)計(jì)得到的微博平臺(tái)特征信息進(jìn)行信息重排,對(duì)重排的微博特征信息信息流采用神經(jīng)網(wǎng)絡(luò)訓(xùn)練方法進(jìn)行主成分特征提取,對(duì)提取的主成分特征采用優(yōu)化的螢火蟲(chóng)算法進(jìn)行特征篩選和自組織監(jiān)督學(xué)習(xí),實(shí)現(xiàn)微博平臺(tái)信息的優(yōu)化檢測(cè)。仿真結(jié)果表明,采用上述方法進(jìn)行微博信息準(zhǔn)確檢測(cè)準(zhǔn)確度較高,需要的先驗(yàn)樣本知識(shí)相對(duì)較小,檢測(cè)的可靠度得到保證。
[Abstract]:The research of Weibo platform feature information optimization detection can efficiently retrieve the information needed in massive Weibo information. In the process of optimizing the detection of feature information, we need to rearrange the information and extract the principal component feature to complete the detection of feature information. The traditional method combines the correlation test to process and analyze the sample information flow, but neglects the feature extraction of the principal components of the information, which leads to the low detection accuracy. Based on firefly optimization neural network, the simulation of Weibo platform feature information optimization detection is proposed. The autoregressive moving average model is used to rearrange the feature information of the Weibo platform, and the neural network training method is used to extract the principal component feature of the rearranged Weibo feature information flow. The extracted principal component features are selected by the optimized firefly algorithm and self-organized supervised learning to realize the optimal detection of Weibo platform information. The simulation results show that the accuracy of accurate detection of Weibo information by using the above method is high, the knowledge of prior samples is relatively small, and the reliability of detection is guaranteed.
【作者單位】： 常州大學(xué)信息科學(xué)與工程學(xué)院;
【基金】：國(guó)家自然科學(xué)基金項(xiàng)目(61272367)
【分類號(hào)】：TP18;TP393.09
，

本文編號(hào)：2102563