Scribd
Upload
51 views

Introduction Pattern Recognition

This document provides an introduction to pattern recognition, including basic principles and concepts. It discusses that a pattern refers to any discrete event related to an application. Patterns can be described using a small number of parameters or features to characterize an object. Patterns are represented in multidimensional feature spaces, where each dimension corresponds to a feature. The pattern space of an application includes all possible patterns represented as a matrix of events, with each event being a point in the multidimensional space defined by the features. Distances between patterns in a space can be measured using metrics like Minkowski, City-Block, Euclidean, Chebychev, and Mahalanobis.

Uploaded by

Ayaz
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
51 views

Introduction Pattern Recognition

This document provides an introduction to pattern recognition, including basic principles and concepts. It discusses that a pattern refers to any discrete event related to an application. Patterns can be described using a small number of parameters or features to characterize an object. Patterns are represented in multidimensional feature spaces, where each dimension corresponds to a feature. The pattern space of an application includes all possible patterns represented as a matrix of events, with each event being a point in the multidimensional space defined by the features. Distances between patterns in a space can be measured using metrics like Minkowski, City-Block, Euclidean, Chebychev, and Mahalanobis.

Uploaded by

Ayaz
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 10

Introduction PatternRecognition

ACOE402NeuralNetworksandFuzzysystems EfthyvoulosC.Kyriacou(PhD)Assis.Prof. ComputerScienceandEngineeringDepartment

P tt Recognition Pattern R iti Basic B i P Principles i i l


Patterniseverydiscreteeventrelatedtoan application li ti Forexampleallpossiblepatternsofenglish alphabetlettersdefineapatternspace

Differententitiesthatrefertothesamepattern

D Describing ibi apattern tt


Describingapattern
Theuseofasmallnumberofparametersinorder todescribeanobject

Processingofsignalsrelatedtopatterns Syntacticaldescriptionofpatternsbydefining primitives Listsofcharacteristics

Describing gpatterns p using g multidimensionalspaces


Use a parameters vector of p dimensions

The numbers could be real, integer or binary depending p g on the p problem we have to describe. They are p features or measurements created through feature extraction techniques
5

P tt space Pattern
Thepatternspaceofanapplicationis expressed dwith ith and d i includes l d amatrix t i of fN events Everyeventisapointinthepdimensional space Whenhavingp3pointscanbe space. graphicallyrepresented

ExampleofaPatternspaceforp=2
7

Description p ofp patternsusing g multidimensionalspaces


Ifandarepatternvectorsthedistances d l din declared i the th applications li ti h haveth thef following ll i properties p p

Di t Distances
Minkowski sclass

City Block

Euclidean

Chebychev

Mahalanobis

10

You might also like