圖計算與推薦系統(tǒng)

Contents..目　　錄
推薦序一
推薦序二
推薦序三
前言
第一篇 圖數(shù)據(jù)與圖模型
第1章 圖數(shù)據(jù)基礎 ··························2
1.1 數(shù)學基礎 ·····································2
1.2 圖的基本知識 ······························4
1.2.1.什么是圖 ·························4
1.2.2.圖中基本元素及定義 ·········5
1.3 圖的表示方法 ····························10
1.3.1.圖的代數(shù)表示 ················11
1.3.2.圖的遍歷 ·······················13
1.4 圖數(shù)據(jù)及圖神經(jīng)網(wǎng)絡 ··················14　
1.4.1.圖數(shù)據(jù)的性質(zhì) ················14
1.4.2.圖數(shù)據(jù)應用 ···················15
1.4.3.圖神經(jīng)網(wǎng)絡的發(fā)展史 ·······16
1.5 本章小結(jié) ···································17
第2章 圖神經(jīng)網(wǎng)絡基礎 ·················18
2.1 神經(jīng)網(wǎng)絡的基本知識 ··················18
2.1.1.神經(jīng)元 ··························19
2.1.2.前饋神經(jīng)網(wǎng)絡 ················22
2.1.3.反向傳播 ·······················23
2.2 卷積神經(jīng)網(wǎng)絡 ····························24
2.2.1.卷積神經(jīng)網(wǎng)絡基本概念
和特點 ··························25
2.2.2.卷積神經(jīng)網(wǎng)絡模型 ··········29
2.3 循環(huán)神經(jīng)網(wǎng)絡 ····························30
2.3.1.循環(huán)神經(jīng)網(wǎng)絡結(jié)構(gòu)和
特點 ·····························31
2.3.2.循環(huán)神經(jīng)網(wǎng)絡模型 ··········35
2.4 圖神經(jīng)網(wǎng)絡 ································36
2.4.1.圖神經(jīng)網(wǎng)絡綜述 ·············36
2.4.2.卷積圖神經(jīng)網(wǎng)絡 ·············41
2.4.3.循環(huán)圖神經(jīng)網(wǎng)絡 ·············42
2.5 本章小結(jié) ···································44
第3章 知識圖譜基礎 ·····················46
3.1 知識圖譜的定義和模型 ···············46
3.1.1.知識圖譜定義 ················47
3.1.2.知識圖譜嵌入 ················48
3.1.3.距離變換模型 ················51
3.1.4.語義匹配模型 ················53
3.2 知識圖譜上的神經(jīng)網(wǎng)絡 ···············55
3.2.1.關(guān)系圖卷積網(wǎng)絡 ·············55
3.2.2.知識圖譜與注意力模型 ·····55
3.3 本章小結(jié) ···································59
第二篇 推薦系統(tǒng)
第4章 推薦系統(tǒng)架構(gòu) ·····················62
4.1 推薦系統(tǒng)的邏輯架構(gòu) ··················62
4.2 推薦系統(tǒng)的技術(shù)架構(gòu) ··················67
4.3 推薦系統(tǒng)的數(shù)據(jù)和模型部分 ········69
4.3.1.推薦系統(tǒng)中的數(shù)據(jù)平臺
建設 ·····························69
4.3.2.推薦系統(tǒng)中的數(shù)據(jù)挖掘
方法 ·····························73
4.3.3.推薦系統(tǒng)模型 ················76
4.4 推薦系統(tǒng)的評估 ·························81
4.4.1.推薦系統(tǒng)的評估實驗
方法 ·····························89
4.4.2.離線評估 ·······················89
4.4.3.在線評估 ·······················92
4.5 基于GNN的推薦系統(tǒng)架構(gòu) ·········94
4.6 本章小結(jié) ···································96
第5章　基于GNN的推薦系統(tǒng)構(gòu)
建基礎 ·······························97
5.1 關(guān)于嵌入 ···································97
5.2 Word2Vec ·································102
5.2.1.哈夫曼樹與哈夫曼編