文章目录
Research contents of AI
1. Knowledge representation （知识表示）
- It is the formalization or modeling of human knowledge.
知识表示就是将人类知识形式化或者模型化。 - Knowledge representation method（知识表示方法）：
- Symbolic representation is to express knowledge using Symbols with a specific meaning in different ways and in different order. For example, first order logic, production, etc. 符号表示法：用各种包含具体含义的符号，以各种不同的方式和顺序组合起来表示知识的一类方法。例如，一阶逻辑、产生式等。
- Join mechanism representation: connect all kinds of physical
objects in different ways and in different order, and transfer
and process all kinds of information with specific meaning
between them, so as to express the related concepts and
knowledge. For example, neural network and so on.
连接机制表示法：把各种物理对象以不同的方式及顺序连接起来，并在其间互相传递及加工各种包含具体意义的信息，以此来表示相关的概念及知识。例如，神经网络等。
2. Machine perception （机器感知）
Machine perception is to make a machine (a computer) have a perception ability
similar to that of a human being.
It mainly refers to Machine vision (machine vision) and machine hearing.
机器感知：是使机器（计算机）具有类似于人的感知能力。以机器视觉与机器听
觉为主。
3. Machine thinking (机器思维)
Machine thinking is to process purposefully the perceived external information
and all kinds of work information within the machine.
机器思维：是指对通过感知得来的外部信息及机器内部的各种工作信息进行有目
的的处理。
4. Machine Learning（机器学习）
Machine learning is to study how to make a computer have a learning ability similar to human beings, so that it can acquire knowledge automatically through learning.
机器学习：研究如何使计算机具有类似于人的学习能力，使它能通过学习自
动地获取知识。
In 1958, Rosenblatt successfully developed the perceptron.
1958年，Rosenblatt研制成功了感知机。
5. Machine Behavior（机器行为）
- Corresponding to human‘s behavioral ability, machine behavior mainly refers to the ability of computer expression, that is, “speak”, “write”, “draw” etc.
与人的行为能力相对应，机器行为主要是指计算机的表达能力，即“说”、“写”、“画”等能力。 - For intelligent robot, it should also have the function of human limbs, that is,
“walk”, “take things”, “operate” and so on.对于智能机器人，它还应具有人的四肢功能，即“走”、“取物”、“操作”等。
Applications of Artificial Intelligence
Typical Problems to Which AI is Applied
Computer vision
Image processing
VR, AR and MR
Pattern recognition
Intelligent Diagnosis
Game theory and Strategic planning
Game AI and Gamebot
Machine Translation
Natural language processing, and Chatbot
Nonlinear control, and Robotics
Typical Problems to Which AI is Applied
Artificial life
Automated reasoning
Automation
Biological computing
Concept mining
Data mining
Knowledge representation
Semantic Web
E-mail spam filtering
Litigation
Typical Problems to Which AI is Applied
Robotics
Behavior-based robotics
Cognition
Cybernetics
Development robotics
Evolutionary robotics
Hybrid intelligent system
Intelligent agent
Intelligent Control
AI发展现状*
Four categories of AI goals:
- Humanly——to measure success in terms of fidelity to human performance.
类人地：以对人类表现的逼真度来衡量。 - Rationally——to measure against an ideal performance measure.
理性地：用理想的性能表现来衡量。
A system is rational if it does the right thing, given what it knows.
一个系统如果对已知的知识做出正确的动作，则被称为理性。 - Acting humanly
Kurzweil,1990: To perform functions that require intelligence performed by people.完成需要人类智能所能完成的功能。
Rich and Knight,1991: To make computers do things at which, at the moment, people are better.使计算机去做此时此地人类才能做好的事情。 - Acting rationally
Poole et al.,1998: Computational Intelligence is the study to design intelligent agents.计算智能是研究如何设计智能体。
Nilsson,1998: AI is concerned with intelligent behavior in artifacts.
AI是关注于用人工手段去实现智能行为。 - Thinking humanly
Bellman,1978: The automation of activities that we associate with human thinking …我们与人类思维相关活动的自动化 …
Haugeland,1985: The new effort to make computers think … machines with minds …新的努力使计算机思考 … 机器具有智力 … - Thinking rationally
Charniak and McDermott,1985: The study of mental faculties through the use of computational models.
通过使用计算模型进行心智能力的研究。
Winston,1992: To make computer possible to perceive, reason, and act.
使计算机能够感知、推理、以及动作。
Eight Foundations of AI:
Philosophy, Mathematics, Economics, Neuroscience, Psychology, Computer science, Control theory and cybernetics, and Linguistics.
8个基础学科包括：哲学、数学、经济学、神经科学、心理学、计算机工程、控制理论和控制论和语言学。
Three types of AI:
Weak AI
Also called Artificial Narrow Intelligence (ANI).
弱人工智能：也被称为人工狭义智能(ANI)。
It is non-sentient AI that is focused on one narrow task (just a specific problem).
它是无意识的AI，专注于一个具体的任务（仅针对一个特定的问题）。
Strong AI
Also called Artificial General Intelligence (AGI).
强人工智能：也被称为人工广义智能 (AGI)。
It means a machine with the ability to apply intelligence to any problem. It is a primary goal of artificial intelligence research.
意味着机器具有将智能用于处理任何问题的能力。它是人工智能研究的主要目标。
Super AI
Also called Artificial Super Intelligence (ASI).
超人工智能：亦称人工超级智能(ASI)。
It is a hypothetical agent that possesses intelligence far surpassing that of the brightest and most gifted human minds.
是一个假定的智能体，拥有远远超过聪明和最有天赋的人类大脑的智能。
Also refer to a property of problem-solving systems, e.g., super intelligent language translators or engineering assistants.
也指的是问题求解系统的特性，例如，超级智能语言翻译器或工程助理。
