本文關(guān)鍵詞： 神經(jīng)網(wǎng)絡(luò) 指數(shù)平滑法 ARIMA模型 音樂 預(yù)測(cè)　出處：《西南交通大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：音樂的流行趨勢(shì)可以根據(jù)當(dāng)前的流行藝人表現(xiàn)出來(lái),因此對(duì)音樂流行趨勢(shì)的預(yù)測(cè)也就是對(duì)哪些音樂藝人即將成為未來(lái)一段時(shí)間內(nèi)的流行藝人的預(yù)測(cè)。而判斷某個(gè)藝人是否是流行藝人則可以根據(jù)該藝人最近一段時(shí)間里的音樂試聽量來(lái)判斷。通過(guò)統(tǒng)計(jì)分析用戶對(duì)音樂的操作(試聽、下載、收藏)記錄,預(yù)測(cè)出藝人在下一階段內(nèi)的音樂試聽量,從而可以判斷出哪些藝人在未來(lái)一段時(shí)間內(nèi)音樂試聽量最高,這些藝人即代表著未來(lái)一段時(shí)間內(nèi)的音樂流行趨勢(shì)。本文通過(guò)統(tǒng)計(jì)分析電子音樂平臺(tái)產(chǎn)生的用戶試聽、下載、收藏歌曲的行為記錄,結(jié)合二次指數(shù)平滑法、自回歸移動(dòng)平均模型以及BP神經(jīng)網(wǎng)絡(luò)模型對(duì)藝人歌曲試聽量進(jìn)行了預(yù)測(cè),同時(shí)設(shè)計(jì)并實(shí)現(xiàn)了基于BP神經(jīng)網(wǎng)絡(luò)算法的音樂流行趨勢(shì)預(yù)測(cè)系統(tǒng)。本文的主要研究工作如下:1.通過(guò)閱讀大量國(guó)內(nèi)外文獻(xiàn),研究了國(guó)內(nèi)外音樂試聽量預(yù)測(cè)的研究現(xiàn)狀、神經(jīng)網(wǎng)絡(luò)算法研究的現(xiàn)狀、該算法的特點(diǎn)以及在多個(gè)應(yīng)用領(lǐng)域中的使用情況。重點(diǎn)研究了BP神經(jīng)網(wǎng)絡(luò)算法的應(yīng)用,并對(duì)電子音樂平臺(tái)上產(chǎn)生的基礎(chǔ)數(shù)據(jù)進(jìn)行統(tǒng)計(jì)分析,尋找影響藝人音樂試聽量的主要因素,最后使用BP神經(jīng)網(wǎng)絡(luò)算法對(duì)藝人在接下來(lái)一個(gè)月內(nèi)每天的音樂試聽總量進(jìn)行了預(yù)測(cè)。2.在使用BP神經(jīng)網(wǎng)絡(luò)算法對(duì)藝人音樂試聽量進(jìn)行預(yù)測(cè)的同時(shí),使用了二次指數(shù)平滑法、自回歸移動(dòng)平均模型對(duì)藝人音樂試聽量進(jìn)行預(yù)測(cè),最后對(duì)比三種預(yù)測(cè)方法的預(yù)測(cè)結(jié)果。3.設(shè)計(jì)并實(shí)現(xiàn)了基于BP神經(jīng)網(wǎng)絡(luò)算法的音樂流行趨勢(shì)預(yù)測(cè)系統(tǒng)。該系統(tǒng)是通過(guò)J2EE平臺(tái),結(jié)合web開發(fā)技術(shù)、數(shù)據(jù)庫(kù)技術(shù)、數(shù)據(jù)挖掘技術(shù)進(jìn)行開發(fā)的。可以讓不懂二次指數(shù)平滑法、自回歸移動(dòng)平均模型、BP神經(jīng)網(wǎng)絡(luò)算法等預(yù)測(cè)算法的工作人員也能通過(guò)操此系統(tǒng)預(yù)測(cè)藝人在下一階段中的音樂試聽量,從而判斷出哪些藝人即將代表下一階段的音樂流行趨勢(shì)。最后總結(jié)了本文的研究?jī)?nèi)容,并對(duì)下一步的工作作出展望。
[Abstract]:The pop trend of music can be expressed according to the current pop artists. Therefore, the prediction of pop trends is a prediction of which musical artists will become pop artists for some time to come. And judging whether an artist is a pop artist can be based on the latest period of time. Through the statistical analysis of the user's operation of the music (listen to, listen to, Download, collect) records to predict the amount of music auditions that artists will be listening to in the next stage, so as to determine which artists will have the highest amount of music auditions over the next period of time. These artists represent the trend of music popularity for a period of time in the future. This paper, through statistical analysis of the users' listening, downloading and collecting songs generated by the electronic music platform, combines the quadratic exponential smoothing method. The autoregressive moving average model and BP neural network model were used to predict the audition quantity of artist songs. The main research work of this paper is as follows: 1. By reading a large number of domestic and foreign literature, the research status of music audition prediction at home and abroad is studied. The present situation of neural network algorithm research, the characteristics of the algorithm and its application in many application fields are discussed. The application of BP neural network algorithm is studied, and the basic data generated on the electronic music platform are analyzed statistically. Look for the main factors that affect the amount of music auditions by artists, Finally, BP neural network algorithm is used to predict the total amount of music auditions of artists in the next month. 2. While using BP neural network algorithm to predict the amount of music auditions of artists, the quadratic exponential smoothing method is used. The Auto-regressive moving average model is used to predict the audition quantity of entertainers. Finally, the prediction results of three prediction methods are compared. 3. A music trend prediction system based on BP neural network algorithm is designed and implemented. The system is based on J2EE platform. Combined with web development technology, database technology, data mining technology to develop. Can not understand the quadratic exponential smoothing method, Staff members of prediction algorithms such as the autoregressive moving average model and BP neural network algorithm can also use this system to predict the music auditions of artists in the next stage. In order to judge which artists will represent the next stage of music trends. Finally, this paper summarizes the content of the study, and makes a prospect for the next work.
【學(xué)位授予單位】：西南交通大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP311.52;TP183

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 章毅;郭泉;王建勇;;大數(shù)據(jù)分析的神經(jīng)網(wǎng)絡(luò)方法[J];工程科學(xué)與技術(shù);2017年01期

2 羅偉偉;;時(shí)間序列分析在金融中的應(yīng)用[J];商;2016年30期

3 王睿;漆泰岳;馮劍;雷波;李延;;基于遺傳算法的BP神經(jīng)網(wǎng)絡(luò)隧道施工參數(shù)正反演分析與應(yīng)用[J];鐵道學(xué)報(bào);2016年04期

4 吳海波;劉銀;石S，

本文編號(hào)：1547618