【AI】Course1 Introduction of Arificial Intelligence

What is the goal to build artificial intelligence?
Build an intelligent machine with mathematics and computing techniques to solve complex problems.

Course 1 Intro of AI

The era of big data

In the era of big data, we are embracing the 4th revolution of industry.
Industry 1.0: Mechanization
Industry 2.0: Electrification
Industry 3.0: Informatization
Industry 4.0: Intellectualization

What is Artificial Intelligence?

John McCarthy: “Any task performed by a program or a machine that, if a human carried out the same activity, we would say the human had to apply intelligence to accomplish the task.”

AI: Engineering or Science

Engineering:

1. Build intelligent systems to solve problems in the world.
2. Understanding mechanisms, algorithms, representations for building intelligent systems.

Science:

1. Understanding nature of intelligence.
2. Implementing models of intelligence to evalute and understand.
3. Exploring consequences of different algorithms and representations.

Turing test

Developed by Alan Turing in 1950, is a test of a machine’s ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human.

Expert System

Knowledge from an expert to built the Knowledge Base of the Expert System. So we need to know how to represent the knowledge.
So the non-expert user can use the User Interface, which usually are computers, to query the questions,
Then the user interface transfer the ques to the interence engine, which the engine try to match what is contain in the knowledge base, and give the answer/advice to the user.

What driven ML?

1) Better models
With more variables to fit the data. From Rule-based -> Statistical -> Deep Learning.
2) Better Computing resource
More CPU, GPU, RAM.
3) Also importantly, more data.

What is Big Data-4V

1) Volume
Huge amount of data.
2) Velocity
Speed to create new data.
3) Variety
Plenty of type of data, including structured data, text, pictures, videos and so on.
4) Veracity
uncertained data, which maybe incompleteness, deception data …

What is Machine Learning?

Machine Learning is the field of study that gives the computer the ability to learn without being explicitly programmed.

How to make Machine learn?

First: Data collection.
Second: Feature Extration
Third: Feature Selection
Forth: Make Models.

The difference between Unsupervised learning and Supervised Learning

Label data.
supervised learning use the data with label.
superviesd learning: dataset with labels:((x⁽¹⁾,y⁽¹⁾),…,(x^(m),y^(m))
unsupervised learning: dataset with no labels: (x⁽¹⁾,…,x^(k))

Regression and Classfication

Regression: If y∈R is a continuous variable.
Classification: If the label is a discrete variable.

Reinforcement Learning

Agent and environement interact at discrete time steps:
Observes state
Produces actions
Gets resulting reward
And Produces the next state