本书是面向高等学校各专业学生的基础性通识教材，旨在引领学生系统了解人工智能的基础知识、基本原理和典型方法。内容包括人工智能基础知识和大模型应用两大方面，涵盖人工智能的基本概念、经典人工智能模型与算法、机器学习和深度学习的基本原理和典型算法、生成式人工智能大模型的原理与应用方法、人工智能与社会的关系及人工智能的行业应用等。 通过学习本书，读者不仅能够了解当前人工智能技术的发展情况、主要研究领域和主流技术方法，还能初步掌握如何运用大模型解决学习、工作和生活中的实际问题。 另外，本书配套提供教学视频、课程标准、教学大纲、教学课件、实验项目和习题等资源，方便教师的备课和授课。 本书可作为高等学校各专业的公共课教材，也适用于所有希望了解人工智能技术的广大社会读者。

胡丽莹（1979-），女，汉族，博士，浙江丽水人。福建师范大学计算机与网络空间安全学院副教授，公共计算机教研室负责人。2001年毕业于湘潭大学信息与计算科学专业，获理学学士学位；2004年毕业于厦门大学计算数学专业，获理学硕士学位；2016年获福建师范大学计算数学专业优化计算方向理学博士学位。目前主要从事机器学习、数据挖掘、模式识别等方面的理论及应用研究，具体研究内容有非负矩阵分解、核方法等的优化算法及其在人脸识别、图像压缩及社区发现等领域的应用。

第1章 绪 论 ·········································································.1
1.1 什么是人工智能 ·····································································.1
1.1.1 人工智能的定义 ··················································································.2
1.1.2 人工智能的五个等级 ············································································.2
1.2 人工智能的发展历程 ·······························································.5
1.2.1 人工智能的诞生 ··················································································.6
1.2.2 人工智能发展的六个阶段 ······································································.7
1.2.3 人工智能在我国的发展历程 ···································································.8
1.3 人工智能的主要研究内容 ························································.10
1.3.1 认知科学 ··························································································.10
1.3.2 知识表示、推理和应用 ········································································.10
1.3.3 机器感知、机器思维、机器学习和机器行为 ············································.11
1.4 人工智能的三大学派 ······························································.12
1.4.1 符号主义学派 ····················································································.12
1.4.2 连接主义学派 ····················································································.13
1.4.3 行为主义学派 ····················································································.13
1.5 人工智能的未来发展趋势 ························································.14
1.5.1 技术演进：向通用智能与自主化跃迁 ·····················································.15
1.5.2 应用深化：从效率工具到产业变革引擎 ··················································.16
1.5.3 挑战应对：平衡创新与风险 ··································································.17
1.5.4 未来图景：人机协同与文明重构 ····························································.18
1.6 人工智能的开发环境 ······························································.19
1.6.1 Python 编程环境 ················································································.19
1.6.2 常用的深度学习框架及安装 ··································································.20/
第2章 经典的人工智能方法 ···················································.25
2.1 知识表示与推理：智能系统的认知基石 ······································.25
2.1.1 逻辑表示法 ·······················································································.27
2.1.2 产生式规则 ·······················································································.30
2.1.3 框架表示法 ·······················································································.32
2.1.4 知识图谱 ··························································································.34
2.1.5 不确定性知识的表示 ···········································································.37
2.2 问题求解与搜索策略 ······························································.38
2.2.1 状态空间搜索：一种表示搜索对象的方法 ···············································.38
2.2.2 搜索策略 ··························································································.40
2.2.3 求解函数最优解的梯度下降法 ·······························································.47
2.3 进化计算与群体智能 ······························································.48
2.3.1 遗传算法：模拟生物进化过程的算法 ·····················································.48
2.3.2 粒子群算法：模拟鸟群行为的群体智能算法 ············································.52
2.3.3 蚁群算法：模拟蚁群行为的群体智能算法 ···············································.55
2.3.4 计算智能算法带给人类的启示 ·······························································.58
第3章 机器学习 ··································································.60
3.1 机器学习：让机器拥有学习能力 ···············································.60
3.1.1 引子：从人类学习到机器学习 ·······························································.60
3.1.2 机器学习的核心概念与发展历程 ····························································.64
3.1.3 机器学习的三大基石：数据、模型、算法 ···············································.66
3.1.4 机器学习的基本流程：从数据到智能决策 ···············································.67
3.1.5 常见的机器学习算法：人工智能的工具箱 ···············································.70
3.2 强化学习：通过反馈学习最优策略 ············································.82
3.2.1 强化学习的基本概念：在与环境交互之中进行学习 ······································.83
3.2.2 强化学习的核心元素 ···········································································.85
3.2.3 强化学习的经典案例 ···········································································.86
第4章 人工神经网络与深度学习 ·············································.91
4.1 从生物神经元到人工神经元 ·····················································.91
4.1.1 生物神经元：大脑的“信号快递站” ·····················································.91
4.1.2 人工神经元 ·······················································································.93
4.1.3 早期人工神经元模型 ···········································································.95
4.1.4 生物启发与人工实现的核心差异 ····························································.96
4.2 人工神经网络的基本结构 ························································.96
4.2.1 激活函数 ··························································································.96
4.2.2 层级结构 ··························································································.99
4.2.3 全连接层 ························································································.101
4.3 人工神经网络的训练过程 ······················································.102
4.3.1 训练流程 ························································································.102
4.3.2 优化技巧 ························································································.107
4.4 深度学习的基本原理 ····························································.108
4.5 深度学习的典型模型 ····························································.109
4.5.1 残差神经网络（ResNet） ············································.109
4.5.2 Transformer ·····················································································.110
4.5.3 生成对抗网络（GAN） ·····································································.111
4.6 深度学习在计算机视觉中的应用 ·············································.112
4.7 自然语言处理与大模型：让机器听懂人类语言 ···························.115
第5章 大模型的应用 ···························································.118
5.1 生成式人工智能 ··································································.118
5.1.1 生成式人工智能应用示例与实操 ··························································.119
5.1.2 国内人工智能生成内容版权规范说明 ···················································.122
5.2 大模型的训练：从数据到智能的进化 ·······································.122
5.2.1 大模型训练过程 ···············································································.122
5.2.2 国内大模型的轻量化训练突破 ·····························································.124
5.2.3 基于LLaMA Factory 的文旅大模型微调实践 ··········································.125
5.3 国外主流大模型：从GPT 到Gemini 的技术盛宴 ························.131
5.4 与大模型沟通的技巧：提示词工程 ··········································.133
5.4.1 提示词工程技巧 ···············································································.133
5.4.2 教育场景提示词工程实操案例 ·····························································.136
5.5 大模型提高办公效率 ····························································.138
5.6 大模型助力科研 ··································································.141
5.7 用大模型进行艺术创作 ·························································.143
5.8 Agent 的原理和应用 ·····························································.148
5.8.1 Agent 的核心原理：像人一样“做事” ··············································.148
5.8.2 MCP 经典案例解析 ···········································································.150
5.8.3 Coze 平台实操教程 ···········································································.153
5.8.4 Agent 的典型应用场景 ···································································.155
5.8.5 Agent 应用的核心特点与未来趋势 ·······················································.156
第6章 人工智能与社会 ·······················································.159
6.1 什么是人工智能伦理 ··································································.159
6.1.1 人工智能伦理的概念 ·········································································.159
6.1.2 人工智能伦理的发展历程 ···································································.160
6.1.3 人工智能伦理的挑战 ·········································································.162
6.2 人工智能安全风险 ·······························································.167
6.3 人工智能伦理治理策略 ·························································.171
6.4 人工智能与社会发展 ····························································.175
6.5 人工智能与创新 ··································································.178
第7章 人工智能的行业应用 ·················································.184
7.1 具身智能 ···········································································.184
7.1.1 案例：浙大开学的机器人搬运工 ··························································.184
7.1.2 具身智能：从物理交互到通用智能的新范式 ··········································.185
7.2 自动驾驶 ···········································································.188
7.2.1 案例：上海奉贤的各类自动驾驶车辆 ···················································.188
7.2.2 自动驾驶：继蒸汽机、内燃机之后的第三次移动 ····································.191
7.3 智能教育 ···········································································.193
7.3.1 案例：“码上”（Mashon）——大模型赋能的智能编程教学应用平台 ·············.193
7.3.2 智能教育：重构教与学的新生态 ··························································.194
7.4 智能医疗 ···········································································.199
7.4.1 案例：智能医疗的临床诊疗辅助作用显著 ·············································.199
7.4.2 智能医疗：面临的核心机遇与主要挑战 ················································.202
7.5 智能家居 ···········································································.203
7.5.1 案例：数字养老社区的智能家居应用 ···················································.203
7.5.2 智能家居：技术融合与生活重塑 ··························································.205
7.6 智能制造 ···········································································.206
7.6.1 案例：智能工厂 ···············································································.206
7.6.2 智能制造：内涵、实践与国家战略·······················································.210
7.7 智能金融 ···········································································.211
7.7.1 案例：校园助学贷款里的“人工智能速通”与“教育公平” ·····················.212
7.7.2 智能金融：人工智能赋能下的风险防控、效率革新与普惠升级 ··················.213
7.8 智能营销：从“猜用户”到“懂用户”——AI 驱动的营销决策新范式
·······························································································.218
7.8.1 为什么传统营销“猜不准”成了常态 ···················································.219
7.8.2 AI 破局：从“数据堆砌”到“读懂需求”的核心 ···································.221
7.8.3 新范式成型：AI 如何重塑营销决策的底层逻辑 ······································.225
7.8.4 平衡与未来：智能营销的“边界”与“可能性” ····································.227
7.9 智慧农业 ···········································································.230
7.9.1 案例：河南小麦基地的人工智能种粮革命
——从“看天吃饭”到“知天而作” ····················································.230
7.9.2 智慧农业：人工智能赋能下的粮食安全守护、
农业效率革新与乡村振兴助力 ·····························································.231