Ch1 Brief Introduction

Posted 2020-07-07

I.Pattern Recognition System

Two processes

Recognition Process:  data acquisition and pretreatment→feature generation→feature extraction and selection→recognition and classification→result　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　

Training Process：　  data acquisition and pretreatment→feature generation→feature extraction and selection→classifiter design（→recognition and classification→result）

II.Pattern Recognition Method

1.Supervised Learning and Unsupervised Learning

Supervised Learning:it need to know the category of every samples.

Unsupervised Leearning:it don‘t need the category of any sample and even the number of samples.

2.Identification model and Production model

The identification model consists of linear and nonlinear.

Identification model:it think that the samples which belong to different category are situated at different regions.

production model:The schema is considered as a random vector in feature space，and every different schema appears in space by probability.