Scribd
Upload
26 views13 pages

Data Compression Using Singular Value Decomposition

The document discusses Singular Value Decomposition (SVD) as a fundamental technique in linear algebra used for data compression and dimensionality reduction. It outlines the process of decomposing a matrix into its constituent components and approximating the original matrix through reduced dimensions. The mathematical formulation of SVD is also presented, highlighting the roles of the orthogonal matrices and singular values involved.

Uploaded by

abrar2010abu
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
26 views13 pages

Data Compression Using Singular Value Decomposition

The document discusses Singular Value Decomposition (SVD) as a fundamental technique in linear algebra used for data compression and dimensionality reduction. It outlines the process of decomposing a matrix into its constituent components and approximating the original matrix through reduced dimensions. The mathematical formulation of SVD is also presented, highlighting the roles of the orthogonal matrices and singular values involved.

Uploaded by

abrar2010abu
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
You are on page 1/ 13

uring

Dta enion
Cornpresion
|evD.

Eingul as Value. De compoeitien CsvD)


u Va anoad algeba Heehniq that
rn ue wed Jdo comu data,
hat
üncuding
douen the
b.
matix üNto t! galte.
SVD ú aa tundamertal ontut
in uine ary Valgobra and w utod în
eaning toy atke ueh
mathine
machine u
anodud
au dimnenion atuodu t ion,ion, noise
noue
yedudin and dta compress LOn
Thu çy how eVD WORky in data

I: De composeg the matix: cVD beakt


down'a maiy înto entituent
inclun etot't
ve
etons,agoral
and a eia gonalmatar.
2: ReducL Dimonironauyi- Vcu
A0cuce the dimentim ality )the
docta.
Appyoximatt the mata,SVD
aemallr Nank to appoyîmatu
the. original malix
4. cn
îm .
oiginal îmagervotutnl the.prouut

Torymuladion.
Mathematical

wherc: minrom
A= origlnal mativ
md o
orthogonal matax
U= an mxncolumnJ ane clled the
whose
inglar veetors ith
dagonal mtyix
E= an' mxn on the diagona
ingula vaule!
whach veuentt th mapil hude

V
T
withnpondt
e 2h omponrt
nxn othogonal motix
whase oWs arl the
Bibarula ve ctoYs.
ant
4+3 -
(2-)-
0xI C2-)
-A 2
2
=0 -AI A'
polmamial pufic charact
2
1 2
atain valt.
o} eingula find I.I
A
Jmati.
oyon vnlua
thexpanii
duad. and!a
decompostion
value ainntitas agedlced Find 9
t/=
ale ctons
22
vallesve Cr]x
- 22
eiqen -3
2 t2|= 2
2-1
eiqen find Promhet
ma= 2-I
Ine
To
Va =

Nom alue Vi and Va.


FoY Vi
V
VI=Cl, )
ength=VI2+12= Va
| Vl = Va

VI=

Va

Va = C -1, 1)
Va =

Combîne VI and ve.


V=|

VT=2
lemputs

Ga = Vha= = l
< wiu e th motix whosc el oa
ertieg ae r n
diaqorial
nce odu shoutel lee
=
AAT

LXI+IX| IXo+1XI ox+Ixo


öXo+IX|
kot ox| LXI +oXo

IA-tl-0
1

(2-)i- -A+AJ-Ci-)9
R-4)C -A -)+2)- CI-A)
CQ-CI-a)+A2) +l-I4\)

9+4-34, =o

-3
1 -3

- 3) =0

)= 3,I,D
to A=3

1 D |-3
CIxo) -ax)| c t-(0x-1)|-Ira)

ngh Va4+i2=
moemaise uCalve

atb 4c-0D
a tobtoc o
atoh t 0 c o ) .
cIXo )-Cox)| Ix) -COx)<xo)
-|

-Vo+(2-2 Va
moA mali ge Ua "ol/e
L-/ve

-Cox)| c2)-ctx
(Ixo )-CIx1)Cix)
enqth= Ve D4I412-|
nom

comine Cu,,U2, U3
mow
o<ve -I|/3

-l|V2

VT=

A
A=

Cix)-(ox

. Rank =2
U= m x .

VT kr

V3
2)f6x1)/ o|f2
-I|/2x4|Ve
V36
1l/2
xV3a)v6 Ronk:

You might also like