本文選題：量子機(jī)器學(xué)習(xí) + 量子k-means算法��； 參考：《南京信息工程大學(xué)》2017年碩士論文

【摘要】：量子計(jì)算作為一種新型的計(jì)算模式,已成為解決摩爾律失效問(wèn)題的一個(gè)可能的解決方法。隨著對(duì)量子計(jì)算和機(jī)器學(xué)習(xí)的深入研究,量子機(jī)器學(xué)習(xí)也應(yīng)運(yùn)而生。本文探討和研究量子機(jī)器學(xué)習(xí)中的量子聚類算法(量子k-means算法)和量子圖像識(shí)別。主要研究?jī)?nèi)容如下:(1)基于經(jīng)典的k-means算法,提出了一個(gè)高效的基于距離最小化原則的量子k-means算法。將經(jīng)典k-means算法的部分步驟使用量子算法來(lái)實(shí)現(xiàn),利用量子疊加態(tài)和量子并行計(jì)算的特性來(lái)實(shí)現(xiàn),與經(jīng)典的k-means算法相比帶來(lái)了指數(shù)加速。算法中,為了計(jì)算待分類點(diǎn)與聚類中心之間距離,通過(guò)增加一個(gè)輔助粒子構(gòu)造聚類中心與待分類點(diǎn)的糾纏態(tài),并對(duì)輔助粒子進(jìn)行投影測(cè)量,進(jìn)而依據(jù)測(cè)量結(jié)果計(jì)算出兩點(diǎn)之間距離。算法的目的是將待分類的點(diǎn)按距離最小原則分到相應(yīng)的聚類中。(2)提出了比較兩幅量子圖像相似度的算法,并給出算法的量子線路圖。所提出的比較算法,是在不連接圖像的基礎(chǔ)上,將圖像用量子態(tài)表示,進(jìn)行控制交換(c-Swap)操作,再進(jìn)行量子測(cè)量,根據(jù)測(cè)量結(jié)果判斷兩幅圖像的相似度。(3)將所提的量子相似度比較算法應(yīng)用到量子手勢(shì)識(shí)別中。在經(jīng)典領(lǐng)域中,手勢(shì)識(shí)別的流程比較復(fù)雜。而在量子領(lǐng)域中,無(wú)需提取手勢(shì)的顏色、紋理、特征等步驟,直接可以將手勢(shì)進(jìn)行二值化表示,再根據(jù)(2)中所提的圖像相似度算法來(lái)實(shí)現(xiàn)手勢(shì)識(shí)別。
[Abstract]:As a new computational model, quantum computing has become a possible solution to the problem of molar law failure. With the further study of quantum computing and machine learning, quantum machine learning has emerged as the times require. Quantum clustering algorithm (quantum k-means algorithm) and quantum image recognition in quantum machine learning are discussed and studied in this paper. The main contents are as follows: (1) based on the classical k-means algorithm, an efficient quantum k-means algorithm based on the principle of distance minimization is proposed. Some steps of classical k-means algorithm are implemented by quantum algorithm and quantum superposition state and quantum parallel computation. Compared with classical k-means algorithm, it brings exponential acceleration. In order to calculate the distance between the points to be classified and the centers of clustering, the entangled states between the centers of clustering and the points to be classified are constructed by adding an auxiliary particle, and the projection measurements of the auxiliary particles are carried out. Then the distance between two points is calculated according to the measurement results. The purpose of the algorithm is to divide the points to be classified into the corresponding cluster according to the principle of minimum distance. (2) an algorithm to compare the similarity between two quantum images is proposed, and the quantum circuit diagram of the algorithm is given. The proposed comparison algorithm is based on the disconnection of the image, the image is represented by quantum state, the control switching c-Swap-operation is carried out, and then the quantum measurement is carried out. The proposed quantum similarity comparison algorithm is applied to quantum gesture recognition. In the classical field, the process of gesture recognition is complicated. In the quantum field, without extracting the color, texture and features of the gesture, the gesture can be directly binary representation, and then the gesture recognition can be realized according to the image similarity algorithm proposed in X2).
【學(xué)位授予單位】：南京信息工程大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP391.41;TP18

【參考文獻(xiàn)】

相關(guān)期刊論文 前8條

1 陳漢武;高越;張軍;;量子K-近鄰算法[J];東南大學(xué)學(xué)報(bào)(自然科學(xué)版);2015年04期

2 阮越;陳漢武;劉志昊;張俊;朱皖寧;;量子主成分分析算法[J];計(jì)算機(jī)學(xué)報(bào);2014年03期

3 于一;;K-近鄰法的文本分類算法分析與改進(jìn)[J];火力與指揮控制;2008年04期

4 熊承義;李玉海;;統(tǒng)計(jì)模式識(shí)別及其發(fā)展現(xiàn)狀綜述[J];科技進(jìn)步與對(duì)策;2003年09期

5 張良國(guó),吳江琴,高文,姚鴻勛;基于Hausdorff距離的手勢(shì)識(shí)別[J];中國(guó)圖象圖形學(xué)報(bào);2002年11期

6 李晗,計(jì)時(shí)鳴,張利;馬氏距離不變量用于灰度圖像的識(shí)別[J];浙江工業(yè)大學(xué)學(xué)報(bào);2001年04期

7 任海兵,祝遠(yuǎn)新,徐光,林學(xué),張嘵平;基于視覺(jué)手勢(shì)識(shí)別的研究—綜述[J];電子學(xué)報(bào);2000年02期

8 黃智輝;判別函數(shù)模糊模式識(shí)別法[J];煤田地質(zhì)與勘探;1990年01期

相關(guān)碩士學(xué)位論文 前1條

1 倪時(shí)策;模式識(shí)別的算法加速器關(guān)鍵技術(shù)研究[D];國(guó)防科學(xué)技術(shù)大學(xué);2009年

，

本文編號(hào)：1839409