clAssmate

Date
29)2p4
Page
ASSiqhment 2

1
Dra co the archite eture of DisCete topfield netoork
is
What
funeton of Discrete Hopffel4
petwork
Eplain in deta) discxete Hepfield netcaost
training testing.
A)
Daw Hopfidd er0Y neteoo xk cuith4 olp node
5hat ase the feattu re of koh non sf
leasning alosith m.
G)
Defne tbe follocotng
) Patten asso cictlon
R) PatteYn eCoqnization
approtinaian
) Centol plant

vi) Beam forming

2)hat shoxt note of Self oTqannaton map

8\List Ut feo appli cations of tohnon self organizadlani

map.
9
classify the foux in put cluster sing
DX
GAni2atian map
Caoo)) assume lear ninq 9ate as oS initil

12
map he
o)-Po kahnen Slt orqani2a}jon if
he qiven eucleh
oeightS aTeShocon clasest toUp veo fo

findfhe clusfer of
distance coElght fo cunit yi
Co2, o4) nd the
 clAsSMAte
pate
Page

b)
for i/p vectoy Coi6.o.6) coith learniog Yate
find eoinning cluste uniteits neco soeight.

o.2

J4e shost hote cn Enexqy mini mization tune tioh

r2)hat the hopfield netooxk. DiscusS
is elahan.
betn stabie sta tes of blnaxy hop fleld no e qraded
version of model
13)
Degeibe the Conce pt of vector coth missing element
14> Construct auto associaive nlo to tore the
Nector JSe he jtesave auto associative
netaoTK to test the vecto lto 2
missing elemest
5) USe he Guter podact le to store he vee tor
aSsociative
netcoork. fnd
Dfìnd 0eiq nt matix Cdo not set diagona element Js O

H)
Tet the ector LS)ngCI Ias input
16Eplain Indetai) about k- mean clugtexing alqoithm
1)Ahoat are the appicaions of clusteslg
algorithm
1o)coi te a
shoxt note on winne takes ad learning
oY
Competaive earni nq ue.
19)con Si der he qiven ip pattenPrPa, Pa,P4 P5
ini-ial Loeight axe leashl
Considey
 cAssMAte
Date
Page

ate o
so.S, pextorm he olustesing usingiotnnet takag
all alqoi thmhe i/p pa He n
are pesented in the
followinq Seq uence P4
P, -o. 96)
PaPaPPs Ps
toyo.707L
o.9806
PoO.|g6) Ps-o58)2 6. 707)
o-98o6 70-8 73
0.707
Ps-o.8137
-o 5S12

20)
Explain the basic once pt of machine Jeanìng is
ypes.
List out the few appli catons of machine learnin e
2D

ohat are Jhe thee key elements at machine learning

2Eplein the Conce pt of Sappoat vec tor machìne
23)|Discess the apti mad byper pla ne(29Dis cass the optiml
hypesplane fox as Îineaxty se parahe patten.
O52+) biscuss
the Opimal by pex plane foY non Separable
plan pattern
28)2) hat do you megn lby InDeN poo ductkesnel
o

2724) bì scuss the MercerS 4heoe

m

28))cassid bricfly classify the Suppoxt ve ctor machine

23) Explain the concept of zad negressisnSupport
ne
Machi
3o)
oa)o hat is tbe east megn algorithm toaining
for

eSinq otet
E>Pes form the too toaining Stcp fon nlo
e
coithcsa )
of leasning
 Page

Co.25
limitations algohm.e benefi+ of MS
32)hat are the tohy
eepleaningmefoset
33) Explain the concept of e deep leazning
g4) pave Machine leasnlng
Com

35) Expl ain the diffesent types of deepTeoning

Convolutlon nl Ohat isoCulan
36) Epla) he per atlon
discaete Convolutio,
Convola H6n neural netwozh.
3) aoe aychi tecture
a

G8)
LDite shoxt nete
layex
h) padeling
N) Stided conuoluHon
) Rectified lineas unit RELU)
39)
hat are he fof bidden layer
40)
bat axe the li rm)tatjons of RELU
2Explan about he lioeax Vasiant
DLeaky RellV
) Pad moic Rel
42) Explain the folloinq hon li neam waviant
)a ussian EgOY lineas Unit
) Soft plus O
Smoob
)Exponential ineax tini t
43)
Exdain the Concept ofane layexe Convaluon netnt
44) liefy degcribe the Conneced layyer.
45) Deg
tnibethe complex )mage CkisIfiCatlon 4S ng
CNN