0% found this document useful (0 votes)
39 views10 pages

Confusion Matrix Explained for Classification

Uploaded by

ranaji181223
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
0% found this document useful (0 votes)
39 views10 pages

Confusion Matrix Explained for Classification

Uploaded by

ranaji181223
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/ 10

I

{n*4n

l
Z

A confusian mat:lix rs a ta[ le*hah cqlegoriees


predicEions accoding ld wheJhev they ma)^ch l*te qclual
L- vqlue inlhe dala. Ane ol the table's divnenaians rrdrcufus
I lhe yassible cofepries a{ yediclecl values while the alhev
dimensian indicqks the sawe ftr oclua I values. fr contudion
F- tnqlri x (dyt be cveufud firr a model pvedicliry cny vumber of
cla sses.
tNhen l*e predicled vslre &e bame (s lhe acluql
r.s
value,fhis is q eortr€cf claosiFrca fion. frwect predichons
{all an lhe dlaganal inthe tonfuaion vnatrix.lTte off -diagonol
watvix cells lrtdieale cqs€s u:here the predicted yalue
"l;fters bam fle tcluql vqlue.llese ave incowecL predrct,rns,
L Pe"r{wmance vnel,su?res frr classiFica lian 'madels ure bq.oed
on the couT?fs of predictr0n6 falliry an and of{: lhe diagonal
in lltese lableo.
the most co ynman pevfatm,erce meqouyes &rtsider
*he model's ability"{c discerr 0}x€ clqss versus alt athers.
lhe clqss oF i rateyesf is kro er ds *be pos'f lrve class, ahile
c.ll olhers ore kro*$rr as neg*hve. Nofu lhat llte use of lhe
tezryr inalogy posifrve and rrega|,ve ls ',40b infended fo imply
any vqlue judgernerrff-e.good veysus bad)rt)or does it
rec€osarily ouggesb tha| tl,te aufco rne is preserrf av absent
ti e - birlh defucb versu 6 nc)ne).The chodce of lhe posi#ue
aubcome &)? gven bp azbllrqryros in cases u.:leye q ynodel
is predicfrng ealegorics such cs su)?ny versus rainy ,ar
drg versu.r ml el-c-

r the relaho nship behruae)1 posi*ve class qndlrega#vc"


L._ class prediclrors cqn be s'hou:)1 {s fi 2x L canfuslan rnqlrit,
l*al fabulales *rhefh et yedicJrbrs full tnfa caft.goa'ie.r *
_+__
ltirlsr P-:s

'no yqs

frc
t

fl{curq

fFN
f
1

+h

Leut-ttivla
\, u
A.l o'h,rul n h,elo*l

h,, I LAL L
J

29 154

154 + leOe
+-
t54+i202t5+-29
+.-
5+!_9
l\q+tentf 5 +2s

S tg rrul ile
*-t-

{rnor€ssive
J\'
ir cJecis i*n rnakino.
U
fcr exq]flple I
. dy] e*

ham
f
sen$fLi vi

t, ,

c etll-l r*l

t\ t51
i54 t 33

LAr-
I
t20; + 5
\1

t_
rec,s t6fl =

ryecq ll =
t

/org u

e{l

= l5tl
rnvecis isn 3= O'96&5
154 f5
t54
15,, * 39

&e&r,ree )., s rlis i lf ui {r;-l r,t


ulb *o
* r-s

ti,:'1"0fi

*ime ds
vfumt fu fus& r y*.r/€fy * { w*,!*,1.,; ivt Y {r, lt"in,l 14ne

* ucs)r€.
6

f- -rreqo(x€ =
veeall + pr"eeiar*r
zX TP
{ZXTil + FP} FN

F* mea;r?€ :: = C'90cs

pffic#ce ls to Hec.such mefis{.rres in c*mbinah'*n uxifh


I

engin€€y.r in {he Field


Y

yertiral

Rq{e
Cse,?s
**is ixrt {ep[ *h re e hy p *t]relrc*rl cla ssifi r v;
ihe Ror plal
#re tcp-arghf tCI?"l?ee of
rs
8ilwe

aufu and c-/ fu lse

zpaalrvc
-U Y€srr ll- Most- tenl-uot/d ,"ln q$iflrpxq' .tre Imilar tn

h*lurc,en nevie cl und u.s€le,ss.


-|--

rnnedi,-Fiva v.rfu-ee *.o uI0


n icyruenblan f*r inl-p":,ove[in*J flUi si'c/e6 {,r,rps r
fo- n cndemL-e Jetlez-gmdes

0'8* 0'!)=B
--+-
o"f *0'S=( tSccepluUe lfuir)
U f =} iPocr)
0'5 * 0'6 =

You might also like