Scribd
Upload
11 views7 pages

Support Vector Machine Overview

The document discusses Support Vector Machines (SVM) and their application in classification problems. It explains the concept of hyperplanes and how they are used to separate data points in a high-dimensional space. Additionally, it touches on optimization techniques to find the optimal hyperplane that maximizes the margin between different classes.

Uploaded by

SOORAJ B
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
11 views7 pages

Support Vector Machine Overview

The document discusses Support Vector Machines (SVM) and their application in classification problems. It explains the concept of hyperplanes and how they are used to separate data points in a high-dimensional space. Additionally, it touches on optimization techniques to find the optimal hyperplane that maximizes the margin between different classes.

Uploaded by

SOORAJ B
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

elASSMAte

Date
Page
Smppo Vector Maehne

SVM aa 4mp er vue4 nehne algonfh mat


Cne mp loye e ame caton and
reavessM prslolkm
n elksificatim
Bmt mssiy, i CcseeA

poblems
he ma bahine W urd ng
pers one n t loat divda the draset
Cnb dnnes Sappodree

nppo Veetors :
ptVedtos a e data poun& neresk
Ine hperplame ne pets d a dutn seb

|hak, moVL uwswl alt he porstuon


v na per pame
p e plane2
cnaepr mple Mampk , ov
chstealm tasl unth my fwo feadares, a
hper lame Cmlbe uouahtd
dAs une dar a

niary seperahs omd alhaisis a se a d


clASSMAte
Date.
Page

fo end gh pep lme ?


The ctane bl porp lr ome
4h meAreet data pors enppari Vectos)
2thew

he qoal behoss e a y per plemn

n dntnbimg
CeyveA

Optin pe n4 pep lane


O

arqin pplnme

Contidey a meavy sapevade dat Set


u Vma twrod
ennd+

Cmnded a sepnsong h Peplone HY he

ohs hne rom he tramna


dhe
cns knces t vafina, hperplre H, arrd
dial nce e dondle thie
SnallestI
eeled fhe masqn
Sdlest dislane
he seperafine
ppleie H
Date.
Page.

5Hypeplme
dhe vqt le
heh
hch1h mar m
marnimal maqw
manimar
p l n also edllkd
hper
naen perp lamt)/ optima sepalvna ypep k

s alse
jhe mamum| mavqn pep lant
celled SYM or d eta se
h e d d pevnts hat e closut to fh maeam

mareinjperps are calld pov Veetov


Auma seportin frpme

4RtbefC
SAppot victa MA
e Oat Co wh eamtks

B Ct

AC, 1+2-5 -o
alAsSMALe
Date
Page

DARet "atl 3 enamples


n0em plk abels.
2
2

BC

AC2,

(o9
Mena maiin herpkre
Page

fomalilin o SVMrablem
Madhemat ed

SYMeblem Hnd
peblan
it lneary
Muma ucin heplene

i m uptim2atto boblm
bsuns
G a huo cas dvaiming dtas

s dijinnnr vea veTov

hhere w ovnl VLoo to hpenpkn

bese
*nce hetaining datA meady sehezlble, we hae
hybesknes dht perale e hov slanes d datu,

n e mAiMumL maqb peplkmt he hqprplume


alASSMAte
Date
Page

those hperplane cqualio

sungle Cmdeon

al iN
dnmce bl o hypers nnes

s ne planes w
mAmimu2e fhe dnstance
munml2

have o seke
qan apinal ajperplome , ere

minimiae
Snen A Vectoy md which

o-k-

LF

meb
elasSMAte
Date
Page

agrant
Lsl tge maltiplus stnleL
emdnq1 Ohe locAl Mameima eyne minu
CunelO Snet egwelia -cmstraints

whove - n Vectov that maumi s

iel

r huro elas Tnenlygeperabe dathsek -


W' boms he ovM
(,)_ 4 . , tn
ashere Ahe eq1' are elkr +l o-
Sp l-Fvnd = (1,k2:A)heay manuriss
1T

S p - Conpte

Skp3- Lmpte b lmin d . ) + man (@ J |

KC)-3.7-L

You might also like