Introduction to Computer science basic knowledge of AI

1. Introduction

1 What is AI?

Artificial intelligence is the research of the program system, which can imitate human activities, such as perception, thinking, learning and reaction to some extent.

2 Brief History of artificial intelligence

3 Turing Test: This test presents a definition of a machine with intelligence.

4 Smart Body: A system that intelligently senses the environment, learns from the environment, and interacts with the environment.

Agents can be divided into two categories: software agents and physical agents.

5 programming languages: While some common languages (such as C, C + +, Java) can be used to write intelligent software, the two languages are specifically designed for AI: Lisp and Prolog

2. Knowledge representation

We describe four common ways to represent knowledge:

1 Semantic Web

2 Frames

3 predicate logic

4 Rule-based systems

3. Expert System

One goal of AI is to build an expert system to accomplish tasks that normally require the experience of human experts. They can be used in situations where human experts are scarce, expensive, or unavailable.

Expert system Architecture:

An expert system consists of 7 components: User, user interface, inference machine, knowledge base, fact base, interpretation system and knowledge Base editor.

1 extracting knowledge → knowledge Base

2 extract facts → facts Library

4. Perception

Another goal of AI is to create machines that behave like ordinary humans (trivial systems).

If an agent is to behave like a human, then it should have the ability to perceive.

Ai has initially completed two senses: visual and auditory.

1 image processing

The first part of this goal involves image processing or computer vision, a field of artificial intelligence that deals with object perception.

Steps:

① Edge Detection

② segment:

Methods: Thresholding, partitioning, and merging

③ Find Depth:

Methods: Stereo Vision and motion

④ Find Direction:

Two techniques: light and texture

⑤ Object Recognition

Applications: such as manufacturing

2 language Comprehension

The second part of this goal is the language processing, analysis and translation of natural language.

Steps:

① speech recognition

② Syntax Analysis:

Two tools: A well-defined grammar and lexical analyzer

Semantic analysis of ③

④ Pragmatic Analysis: Used to further clarify the use of sentences and to eliminate ambiguity

Role: Intent and disambiguation

5. Search

In artificial intelligence, one of the techniques of problem solving is searching.

A search can be described as solving an issue using a set of States (scenarios).

Two common ways to search:

1 Brute force search: breadth-first search and depth-first search

2 heuristic Search: Using it we assign a quantitative value to each node as the heuristic value (h value).

6. Neural network

If the agent should behave like a human, then it may need to learn. Several of the methods that have been used have created hope for the future. Most methods use inductive learning and learning from examples.

A common approach is to use neural networks to try to mimic the learning process of the human brain using neural networks.

1 biological neurons

2 Perceptron: An artificial neuron, similar to a single creature.

3 Multi-layer network: Several levels of perceptron can be combined to form a multilayer neural network.

Layer Three: Input layer, hidden layer, output layer

4 applications

Two proven neural networks useful: optical character recognition (OCR) and credit assignment.

This article is from the "11094276" blog, please be sure to keep this source http://11104276.blog.51cto.com/11094276/1734820

Introduction to Computer science basic knowledge of AI