Introduction to Computational Linguistics

更新时间:2023-06-09 03:00:01 阅读量: 实用文档 文档下载

说明:文章内容仅供预览,部分内容可能不全。下载后的文档,内容与下面显示的完全一致。下载之前请确认下面内容是否您想要的,是否完整无缺。

CMC 723S :Inrot t ooCpmutatioalnL igunitscsiFeburry 2a8,2 030

ecLuret4 a Ba:kgcorud fornP obarbliitic NLP srPfoseos: rBonin eJ .DorrAT :Waiiay (Yunn) Chonga

hW (not)y Staisttci fosr LNP? P ro–D iamsbigauton –i ErorrT olrent –aL ernable

aC o– noN alwaysta pporprita –eDiff cilt to udeubg

eiWhtgdeA utmoaatTr/nasducer sS epce hrecgoition: ntsoingr praonucnitaoin elxcon i Aguemntatoin f FSoA:E ch arcai assoscateid ith wap rbaobiilty

Poruncnationi ntewor kfor“ aobut”

oNiy Chasnel

Pnrbobaliiy Dtfeiinitosn E xprieent (tmrail–)Repeat abe prloedcue rwih wellt-dfieen dosspbil eotuomesc Smape splcea Co–plemte se otfout omecs vEent–A yn usbse ot oufctmes ofor sammlep pscea Rand om Vaiarlbe– Unertcani outomec i na riatl

Mor Defeniiitosn Prbaoblitiy– ow likHle ys ii totget apar itcuar oltuoce?m – aReto fgetint ghat outcotme in la ltralsirobabPiilytof dra ing w spaad fremo 5 w2el-lhusflfedp lyinga ardcs:

D siritubtion Pr:baoibltieis asocsaietdw it haech uotocm e ranaomdv araibe cal ntkea– Eca hutcome oash rpboailibt yetbewe 0 andn1 – he Tsm uof allou cote probmbialites is 1i.

C

ndotioial nrPobailitby hWa ti Ps(|A)B ? Fist,rw at ish (PA)?– P(It “sirai nign) ” .=60 Nwow ha tbauot (P|AB?)– (PIt“i sariing”n “|It was learc 10 imnutse aog) ”=. 004P( A B)| P (A, ) BPB)(

A

AB B,Noet: PA(,)B=(PAB| ·) (PB Al)s: PoA,()B= (PB,A)

Indpeneencde Waht i PsA,B( if) andA aBrei ndpeednet?

n (P,BA)P(A=) ·P(B) if AfB, ndiepndene.tP(head,stails =) (hePda) ·sP t(ail) s=.5 ·= .2 5. P(do5ctr,olbu-eyee)s P=doctor) (· (Pble-uyes)e .0= · .1 2= 002 N.ote: PA(|)BP(=A) if AfB in,depneend tAslo:P(B A)=P(|) ifBf ,B Aidepnedennt

Ba

ey TseorehP m (B | A )P (A |B )P()BP ()A wSp tae hodrreof depe dencen oSetimmse esier toa esimatte oen knid o fdpeenendec thant he toehr

Whtad oe thissh avet o do iwt the Noihsy Channle odeM?

(H)l

(O

)(H)PP(|O)HBes HP tO |( ) H P H ) P (O( ilkeilhodo) pirroBes Ht a=rgam PxH(O|) =argaxm H

HoNiysC anhenl Appiel dt Wood Recrognitio nar gmxwaP (|O) w=a gmarw Px(Ow| P(w) ) impSlfyiin asgumptionss p–rounnicaiot sntirgnc roerc –t wordb uodarnisekn onw

robleP:m –ivGne n [i], ywat hi scorret cidtiocarnywor d

?What o dwe een?dn[i:] knee, ean,tn ede ,ew

nW

aht i the smsto ileky lwrd giveo nn[]i? ompCue prior t(P)worW dfeqr(w P()w)newneat nee dkeen26523381 17 641.001

.00103 0.0065 000.024 N w comoupt elikliehoo Pd[n(i|]w)orWd enw nat eenedk eenP (|Ow .3) 652. .1 1.10 P(w0 ).010 00.103 .00506. 00002 4(OP|)P(ww ).00360.00 0068. 000620.000 240

本文来源:https://www.bwwdw.com/article/7m21.html

Top