本文選題：MapReduce + 樹增強(qiáng)型貝葉斯算法�。� 參考：《激光雜志》2015年12期

【摘要】：為了解決大數(shù)據(jù)環(huán)境下數(shù)據(jù)日益增大且響應(yīng)時間要求變短,以及串行貝葉斯分類器效率低且應(yīng)用復(fù)雜度高的問題,提出了基于MapReduce的并行樹增強(qiáng)型貝葉斯算法。本算法使用了弱化了獨立性的樹增強(qiáng)型貝葉斯算法以獲得更高的分類精度,同時為了降低響應(yīng)時間,引入了MapReduce模型,將本算法由串行轉(zhuǎn)為并行,從而提高處理的速度。實驗結(jié)果表明該算法比傳統(tǒng)的樹增強(qiáng)型貝葉斯算法具有更高的算法效率且隨著數(shù)據(jù)節(jié)點的增加,加速比也同步增加。
[Abstract]:In order to solve the problems of increasing data and shorter response time in big data environment, and low efficiency and high application complexity of serial Bayesian classifier, a parallel tree enhanced Bayesian algorithm based on MapReduce is proposed. In order to reduce the response time, the MapReduce model is introduced to transform the algorithm from serial to parallel, so as to improve the processing speed. Experimental results show that the proposed algorithm is more efficient than the traditional tree enhanced Bayesian algorithm and the speedup ratio increases synchronously with the increase of data nodes.
【作者單位】： 貴陽學(xué)院數(shù)學(xué)與信息科學(xué)學(xué)院;南京財經(jīng)大學(xué)管理科學(xué)與工程學(xué)院;
【基金】：貴州省科技廳聯(lián)合基金(LKG[2013]43號)
【分類號】：TP18;TP338.6
，

本文編號：1956750