移動通信大數(shù)據(jù)分析：數(shù)據(jù)挖掘與機器學(xué)習(xí)實戰(zhàn)

定　價：￥99.00

作　者：	[中] 歐陽曄，[中]胡曼恬著
出版社：	清華大學(xué)出版社
叢編項：	新時代·技術(shù)新未來
標(biāo)　簽：	暫缺

購買這本書可以去

ISBN：	9787302541240	出版時間：	2020-11-01	包裝：	平裝
開本：	16	頁數(shù)：	212	字?jǐn)?shù)：

內(nèi)容簡介

　　本書以4G/5G無線技術(shù)、機器學(xué)習(xí)和數(shù)據(jù)挖掘的新研究和新應(yīng)用為基礎(chǔ)，對分析方法和案例進行研究；從工程和社會科學(xué)的角度，提高讀者對行業(yè)的洞察力，提升運營商的運營效益。本書利用機器學(xué)習(xí)和數(shù)據(jù)挖掘技術(shù)，研究移動網(wǎng)絡(luò)中傳統(tǒng)方法無法解決的問題，包括將數(shù)據(jù)科學(xué)與移動網(wǎng)絡(luò)技術(shù)進行完美結(jié)合的方法、解決方案和算法。本書可以作為研究生、本科生、科研人員、移動網(wǎng)絡(luò)工程師、業(yè)務(wù)分析師、算法分析師、軟件開發(fā)工程師等的參考書，具有很強的實踐指導(dǎo)意義，是不可多得的專業(yè)著作。

作者簡介

　　第一作者簡介歐陽曄博士亞信科技首席技術(shù)官、高級副總裁歐陽曄博士目前全面負(fù)責(zé)亞信科技的技術(shù)與產(chǎn)品的研究、開發(fā)與創(chuàng)新工作。加入亞信科技之前，歐陽曄博士曾任職于美國第一大移動通信運營商威瑞森電信（Verizon）集團，擔(dān)任通信人工智能系統(tǒng)部經(jīng)理，是威瑞森電信的Fellow。歐陽曄博士在移動通信領(lǐng)域擁有豐富的研發(fā)與大型團隊管理經(jīng)驗，工作中承擔(dān)過科學(xué)家、研究員、研發(fā)經(jīng)理、大型研發(fā)團隊負(fù)責(zé)人等多個角色。歐陽曄博士專注于移動通信、數(shù)據(jù)科學(xué)與人工智能領(lǐng)域跨學(xué)科研究，致力于5G網(wǎng)絡(luò)智能化、BSS/OSS融合、通信人工智能、網(wǎng)絡(luò)切片、MEC、網(wǎng)絡(luò)體驗感知、網(wǎng)絡(luò)智能優(yōu)化、5G行業(yè)賦能、云網(wǎng)融合等領(lǐng)域的研發(fā)創(chuàng)新與商業(yè)化。

圖書目錄

第1章概述
1.1 電信業(yè)大數(shù)據(jù)分析 ···························1
1.2 電信大數(shù)據(jù)分析的驅(qū)動力 ················2
1.3 大數(shù)據(jù)分析對電信產(chǎn)業(yè)價值鏈的
益處 ··················································3
1.4 電信大數(shù)據(jù)的實現(xiàn)范圍····················4
1.4.1 網(wǎng)絡(luò)分析 ···················································5
1.4.2 用戶與市場分析 ·······································8
1.4.3 創(chuàng)新的商業(yè)模式 ·······································91.5 本書概要 ··········································9
參考文獻 ·················································10
第2章電信分析方法論
2.1 回歸方法 ········································12
2.1.1 線性回歸 ··················································13
2.1.2 非線性回歸 ··············································15
2.1.3 特征選擇 ··················································16
2.2 分類方法 ········································18
2.2.1 邏輯回歸 ··················································18
2.2.2 其他分類方法 ··········································19
2.3 聚類方法 ········································20
2.3.1 K均值聚類 ··············································21
2.3.2 高斯混合模型 ··········································23
2.3.3 其他聚類方法 ··········································24
2.3.4 聚類方法在電信數(shù)據(jù)中的應(yīng)用 ·················25
2.4 預(yù)測方法 ········································25
2.4.1 時間序列分解 ··········································26
2.4.2 指數(shù)平滑模型 ··········································27
2.4.3 ARIMA模型 ············································28
2.5 神經(jīng)網(wǎng)絡(luò)和深度學(xué)習(xí) ·····················29
2.5.1 神經(jīng)網(wǎng)絡(luò) ··················································29
2.5.2 深度學(xué)習(xí) ··················································31
2.6 強化學(xué)習(xí) ········································32
2.6.1 模型和策略 ··············································33
2.6.2 強化學(xué)習(xí)算法 ··········································33
參考文獻 ·················································34

XII
XII

第3章 LTE網(wǎng)絡(luò)性能趨勢分析
3.1 網(wǎng)絡(luò)性能預(yù)測策略 ·························39
3.1.1 直接預(yù)測策略 ··········································39
3.1.2 分析模型 ··················································39
3.2 網(wǎng)絡(luò)資源與性能指標(biāo)之間的關(guān)系 ···40
3.2.1 LTE網(wǎng)絡(luò)KPI與資源之間的關(guān)系 ···········40
3.2.2 回歸模型 ··················································41
3.3 網(wǎng)絡(luò)資源預(yù)測 ·································43
3.3.1 LTE網(wǎng)絡(luò)流量與資源預(yù)測模型 ···············43
3.3.2 預(yù)測網(wǎng)絡(luò)資源 ··········································43
3.4 評估RRC連接建立的應(yīng)用 ············46
3.4.1 數(shù)據(jù)準(zhǔn)備與特征選取 ······························46
3.4.2 LTE KPI與網(wǎng)絡(luò)資源之間的關(guān)系推導(dǎo) ····47
3.4.3 預(yù)測RRC連接建立成功率 ·····················49
參考文獻 ·················································50
第4章熱門設(shè)備就緒和返修率分析
4.1 設(shè)備返修率與設(shè)備就緒的預(yù)測
策略 ················································53
4.2 設(shè)備返修率和就緒預(yù)測模型 ··········54
4.2.1 預(yù)測模型的移動通信服務(wù) ························54
4.2.2 參數(shù)獲取與存儲 ······································55
4.2.3 分析引擎 ··················································56
4.3 實現(xiàn)和結(jié)果 ·····································58
4.3.1 設(shè)備返修率預(yù)測 ······································58
4.3.2 設(shè)備就緒預(yù)測 ··········································62
第5章 VoLTE語音質(zhì)量評估
5.1 應(yīng)用POLQA評估語音質(zhì)量··········68
5.1.1 POLQA標(biāo)準(zhǔn)···········································68
5.1.2 語音質(zhì)量評價中的可擴展性和
可診斷性 ··················································69
5.2 CrowdMi方法論 ····························69
5.2.1 基于RF特征的分類 ·······························70
5.2.2 網(wǎng)絡(luò)指標(biāo)選擇與聚類 ······························70
5.2.3 網(wǎng)絡(luò)指標(biāo)與POLQA評分之間的關(guān)系····70
5.2.4 模型測試 ··················································70
5.3 CrowdMi中的技術(shù)細節(jié) ·················71
5.3.1 記錄分類 ··················································71
5.3.2 網(wǎng)絡(luò)指標(biāo)的選擇 ······································71
5.3.3 聚類 ·························································72
5.3.4 回歸 ·························································73
5.4 CrowdMi原型設(shè)計與試驗 ·············74
5.4.1 客戶端和服務(wù)器架構(gòu) ······························74
5.4.2 測試和結(jié)果 ··············································76
參考文獻 ·················································78

目錄XIII

目錄XIII
第6章移動APP無線資源使用分析
6.1 起因和系統(tǒng)概述 ·····························80
6.1.1 背景和挑戰(zhàn) ··············································80
6.1.2 移動資源管理 ··········································81
6.1.3 系統(tǒng)概述 ··················································82
6.2 AppWiR眾包工具 ··························83
6.3 AppWiR挖掘算法 ··························84
6.3.1 網(wǎng)絡(luò)指標(biāo)的選擇 ······································84
6.3.2 LOESS方法 ············································87
6.3.3 基于時間序列的網(wǎng)絡(luò)資源使用預(yù)測 ·······87
6.4 實現(xiàn)和試驗 ·····································88
6.4.1 數(shù)據(jù)收集與研究 ······································88
6.4.2 結(jié)果和準(zhǔn)確度 ··········································89
參考文獻 ·················································91
第7章電信數(shù)據(jù)的異常檢測
7.1 模型 ················································93
7.1.1 高斯模型 ··················································94
7.1.2 時間依賴的高斯模型 ······························94
7.1.3 高斯混合模型（GMM）·························95
7.1.4 時間依賴的高斯混合模型 ·······················95
7.1.5 高斯概率潛在語義模型（GPLSA）·······95
7.2 模型對比 ········································97
7.2.1 樣本定義 ··················································97
7.2.2 異常識別 ··················································98
7.2.3 時間依賴GMM與GPLSA的對比 ·········997.3 仿真與討論 ···································100
參考文獻 ···············································103
第8章基于大數(shù)據(jù)分析的LTE網(wǎng)絡(luò)自優(yōu)化
8.1 SON（自組織網(wǎng)絡(luò)）···················105
8.2 APP-SON ······································107
8.3 APP-SON架構(gòu) ·····························108
8.4 APP-SON算法 ·····························110
8.4.1 匈牙利算法輔助聚類（HAAC）··········111
8.4.2 單位回歸輔助聚類數(shù)的確定 ·················114
8.4.3 基于DNN的回歸·································114
8.4.4 每個小區(qū)在時序空間的標(biāo)簽組合 ·········116
8.4.5 基于相似性的參數(shù)調(diào)整 ·························1168.5 仿真與討論 ···································117
參考文獻 ···············································122
第9章電信數(shù)據(jù)和市場營銷
9.2.1 數(shù)據(jù)采集和數(shù)據(jù)類型 ····························130
9.1 電信營銷專題 ·······························127
9.2.2 網(wǎng)絡(luò)的提取和管理 ································131
9.2 社交網(wǎng)絡(luò)的總體構(gòu)建 ···················130

9.3 網(wǎng)絡(luò)結(jié)構(gòu)的度量 ···························133
參考文獻 ···············································135

9.4 網(wǎng)絡(luò)中的消費者行為建模 ············134
第10章傳染式客戶流失
10.1 問題引入 ·····································138
10.1.1 流失率問題 ··········································138
10.1.2 社交學(xué)習(xí)和網(wǎng)絡(luò)效應(yīng) ··························139
10.2 網(wǎng)絡(luò)數(shù)據(jù)的處理 ·························141
10.3 動態(tài)模型 ·····································143
10.3.1 模型介紹 ··············································143
10.3.2 模型的定義 ··········································144
10.3.3 自身經(jīng)驗建模、社交學(xué)習(xí)和
社交網(wǎng)絡(luò)效應(yīng) ······································146
10.3.4 模型估計 ··············································148
10.4 結(jié)果 ············································149
參考文獻 ···············································151
第11章基于社交網(wǎng)絡(luò)的精準(zhǔn)營銷
11.1 網(wǎng)絡(luò)效應(yīng)的渠道 ·························158
11.2 社交網(wǎng)絡(luò)數(shù)據(jù)處理 ·····················159
11.3 建模策略問題 ·····························160
11.3.1 線性空間自回歸模式 ···························160
11.3.2 社交網(wǎng)絡(luò)交互模型 ······························162
11.3.3 內(nèi)生同伴效應(yīng) ······································162
11.4 發(fā)現(xiàn)與應(yīng)用 ·································164
11.4.1 結(jié)果的解釋 ··········································164
11.4.2 基于社交網(wǎng)絡(luò)的精準(zhǔn)營銷 ···················165
參考文獻 ···············································168
第12章社交影響和動態(tài)社交網(wǎng)絡(luò)結(jié)構(gòu)
12.1 動態(tài)模型 ·····································17712.1.1 連續(xù)時間馬爾可夫模型假設(shè) ···············17712.1.2 模型估計與識別 ··································17912.1.3 網(wǎng)絡(luò)結(jié)構(gòu)對社交影響的多元分析 ·······18012.2 研究發(fā)現(xiàn)總結(jié) ·····························18112.2.1 隨機行動者動態(tài)網(wǎng)絡(luò)模型的
估計結(jié)果··············································182

12.2.2 元回歸分析結(jié)果 ··································184
12.2.3 策略模擬 ··············································18812.3 結(jié)論 ············································193
參考文獻 ···············································194