PAGE NO.

ACHIE LEARNIN DATE

Uw Vectons Innenloute_þhodud
nvehe o amatun Ligen Value-
Anal) Singuln Value Dednpobstion.
hobablity Diiubutinn _Londution
Joint háobability Dtuhution,
Boses lheohem

Kow Vedon- 1 31
olmn VetoN=_
3
Madun
AhhayA numbêns , Symbals Lan
be ecdanguah/Rauane Ahang ed ly
houid and tolumns
SO matunn with mhous
and A k u m n s Can he deined

(alled
hev tcAB_
Matn hovina4AtAingie
AInle_h0w/Column
ELemenA_o ecton ah Called
Squahe Matrun
umbe of hoy kLolumn
aheequl mxn_ m: n

Rectongulun Madun
Mumbe o ow&Lolumw
ahe not u. An2 mnf-
Matain AA elemena_he 0
>Digona Matun- t Least element
cia qonal elenet in
Aqcae maJHin iA_hon-Zenn_
20A all elementu he 0
Tolentity Madhin Diagona elemed_ae
97h 3
Jnvene of a matrix.
AA1
A- ady(A)
A

A shuld ko nen ngular malna forA b erst

i Alfo.
Defn
1Tf A 1S a nOn Singulb 0alrix 34uaR nmatrix,
fho iS an existence of nxn matrix A
utuch ia Callodinverse matrix of A 2uch that
AAI:A4A

1)0 acw Ken mataixO

)) 90 Cw then malhi 0 oj
e-1 o

MvHipag Hee a malik -

L
_gt Idonhty matix
 To find imcrse of amatnx.
Miroy
A n
A Cofacte
trenspep-
Hdyerd
Adient Ais thr bansppie o the coactet motax of A
Aj(A)=Cojader(A)
Cofactar can bc sbtained as
Ci (-D'T dok(M2
ind the invese A_a matix

A -2
4 2

M 4 -7 -3+-1
-2J LL 2

A -1
t 2J

a A 23 A'2 9-

L1 2 J LAl: (15-2)-236-42) +3(8-3s)

33t2(6) a1X3
46- 81 78
3
 PAGE NO.

DATE

M-33-6
la 1 2 - 1a
1- 6
-21
3 -6 21
2-1

C 33 29
-1212I2 12
6

A 36
-12
- 12 6
-27 12 3

A- -11
2

siopentie
) dat Codj (A)) LAn-
B[AB) 8-'A
( A B) c-B-A-
(A1)-l= A
Conside1 matn Aqy3 4x3 and 3 he Sid
P x T Y)"pT (3x2)

(3x 3) (3 x3) 9x x2

2X2
 Eigen Values &Vecys
-To Hcduce he dimensiom.

Stock i
Snput
eatue if2 Poodit
taun
ML Mede

Case
A:2
3 J 2

Ax

ARten_aulkiplying L3with matuk ue et reulE _

Matiz in difenE duyecuen

AL 2
l2
AxAL2
2 CL

Ko2lkant vecke U thia t gaal &h

cnechom is Som

A-X 3X
 PAGE NO.

OATE

Ax= AX
(A-) X O
Matix m LAAI6
Eigen airs
Let_A EE RMxn
Let R"* wo ay that_a nMZOAD
Vectei VER is an igen veckei of A_
CeOpend1ig to 1gen value a.(a scala) i
AV: AVV
The abee sdatm fells that when mabuk C0inear
transformatton) A applild om V he0 t sales
V byan amount A.
Howto caluuht
A 2 2 3 AV=Ay (A-V:0
A-AI=O
3

A 2- - 2 3

3
(2-La2--3]+2(-1-A-)t3(3-1tA) =0
(2) CA-4) 2(-A-2)+ 3 (2tA) =o
A+ -8 t2A2-2-yt 6 +3A O
5 A t2A2-6 O
AS-5-2/2+6-O
 -L -2A-5A+G

A-SA+6
2A
-6d+6
-
-A- (A-D:o a (A:2(4-l4-:0
A= L-2, 3
whenA3
A -31= - 2
3 VL
V2
3 -4JVa [oJ]
-V-2Vt3V2 O
V Vz_ V-22 Va:0,
V+3V24V3=O)

A
A-1IE 2

3 -2 Lo
-2V2 13V O
V t Va =0 V
V+3V) -2V3 0_
2V2 P 3 O
voter -2V= 2
 AAT4 2
3 V2
L
4V-2Y2+3V : O
V+2V,t V3 O
-2v, +3V= 4
9V2+V3 3
V2/ Va 14/4
V3 3/-
en vece -1|
14/L I4

9P23 ngulnalu eccopLM=

feaure Data d utizn- (man use n M

USyT
 Orthon0Ya ungth1
gmal d uc0 1

efn
TheSingulan Value de cempetion (SVD 00
mabmx A 0adlen mXn is bo achzation
Anto tha pleduct of_he tioe
matnce3 A= USy
uTÍLTO columnu U2 V_ orto nerma
and ha matrixS is diagonal_wth non-nggatVe
n0al eNta es.

Un| VV2 Un

m xn
UER"
VE
IR

HenR_a 20, 232 n20

Ci 3 Calod uingulan yalue A

DuSt Compute

0Find U
The coluns en Bie olthenomal agen
Veck AAT
onpeMal elpsn
Yekon_aAALaneAprodoct)
T h a i0Rulag value( Cn an Hhe
a 9H00 fhe egen values qAAT CAT A)
 DATE

x i n d Sngulan_value decomposiluon ( Svp) g A_

A 2 2

AAT l

2 22f2
2

42
ATA 2 22

AAT-A T: O
8-A4 o
42L

A=lo
VI O -Vt2V2= 0
LA-4IJ 2V-V:0
V-2VO
V 2 d1o VE2V2

V 2Vt3V= O
L4_2-J V 2V+Vh:O
V2 2V

V2

-2/6
 U= olo bigH ona juil
-2
ToAund V matix

=-0-
5-A=S 5-A-S
A=lo

5 S V L o VtV2=O
S V V-V2
Aslo

- divby l2 6 ake
L IJ o neimal_

V
2

Topnd 2
S_o
Lo o

A: 5
6-/15
Eo - O
 PAGEN1D
DATE

P1obabiliH
linear Alaebra- low Can we repreSent data
Probabilty -Hlow can we talk about uncertaints

Basics of Probability

Defn
An experiment whose outcome is uncertán-
EventSubset o Sample pace
Sample Spare
list/Set of all possible oltcomes

0 Tf die is thrown
= 12,3. 4 5, 6.J

T£Acoins oe tossed
0 HH, HI THTIÍ
Le Set a mepresent Sample space and A be
he set Containing the outcomes of an event theen
PCA) mlA no 0elemen ts in A
no of elementS in 2

Ewhen coin is tossed. HT

A-eveot of Occrence of head
A:iH PCA)V2
Ex Die is thrown L2,3,4,5,61
A-event of Occurence of oulcome div L 3
A:13,61 PA): 2
 PACE NG

TE

event ef that ovtcome is evern

B: 2,4,63 P(6) 362
G)Cbe event thal ovlcome is <2.
C:i1,2,3,4,6,63
Pc): 6/6 1 LSure Event

gv) D is event that outcome

D: PCD)- O LImppSsible event

Complement:
A lz¢A
A:63 PCA):2
A L2,45 PCA'):1/6:6-2 = 1-44 -PA

AT -P(AL.
Unon
AUB a | 2 ¬ A or zEB_Or bothi

B:{2,46] c L 25
BDC 12,2,4,5,43.
pBU): 6/% =L

B2C ae mutully_exhaustive event

Iniersection
AnB {axeA and aeB
 PAGE NO.

TE

BnC
PCBNC): 0
>Mutually exclusive event.

Lwoor more_events are 3aid to be mut exhaustive

if
if thei Lniongives the Sample_Space

TwoOY OCCO
morerenceevents are Said to be Mut.exclosie
L the siOPHe_of an eent excludes the
Occurene of other event.s_

alabaa
P(Y P(xn: PYF
2Y P (Ynx)
P(x
X
PStudent Pass exam)
PLStudent Pass &get 2 10/) S/00 /20
(aiven Shident Pa Exam. PCgetting 90°)?

P Pxn )=2o P(Y

P(Ynx) 20
P(X)
Popplyfor for job)= 4
PPapply YjDb_ ven applin) 2.
PC@aeply iven Papplie) Vg
P P o t oppl e n not ppy)2
 PPne)
P8) PP)
P(oG)
P(G) 2

P(nG)) e (PO'):1-P(PG
78 r78) P10) P(10
2/3 42 /
bl6 S

DPOS)= P(P)t P(G) - P( POO)_

12

3t2-1
12 Chee

Consi der Abags Bag 1

56 56

ne bag Dcfod Ba is diaLOn

PBatl g00n)

PG rt)PG):
6
20 20
 PAGENO

DATE

Total Phobability
PCA) P( A, ):PCE)+P ).PlE)t
MACHINE LEARALING ECHNIQUES_AND PeacLICES
CRircd

raining set

BLd

Gresdlsatuo
Not

Bd WLOVg
tistttt

1ypes o ML SYstem
O SuperVisel ( labellad dala)-fed to the alge
a)laviicahm Cdisaote)
)RopaAKabn-CCombinuou)

npavnd nol labkll A)

a)CtC n
)Anemaly_deteduon Rrnovelty_datutuon
)Visualisatuon diinen&uonality A0oetisn
o2A2s Diahon 2ul eanrig