You are on page 1of 35
statistics et is the science ond — analy2ing docto.- collection, oxgant zing cahat is Data: Data is set of facts o¥ information: atm ot a Location — A Bank has opened 0% ® ond tee is ansthey Loction ©: The Bonk cont fe kre chethes it's usefull fo open a Bank ot ashich Is yo kM. From (A) * fos the abe crasnple ae an ose statistics to help Bank pepk to execute better decession: Types of statistics: ® Descriptive : x Th consists of oganising and summer! 2hy the data. > Applying dippewntt operations te take vee daitas s9: Measure of centyal Tendency Measure Dispersion - Histeg tums, Bax chaxt, pie oy tofsentil: using dota, yoo have measured to form conclusion - 5: 2-test- t. test. Sample questions: a» Let's Soy there ae go students in a meacths closs im the univessity, ond ae’ve colkeled the height of the student mm the chss- Tas, 130) 160, (40, 130 140,160, ° 0 pesciptive? 2 chet is the wengt height * ohat is the most common height - Tnfrenitial: a fre the height of the students ta the classy simnibes to hat you Expect in entive celkge - “soxnple anne popy larlion Datla. ej. exit poll O O © OO susvey Agency ail take sample dita and Makes conchsion of rtite data. Types. of Data: avorttitedive : avomerioa! pisexete: Tdicates specfi¢ ahole nurn bers £3: No.of Rank Accounts: No- of childyen a foxmily continous May have any \alue 3 Heights i eights, Ternperalure, speed avalitative: caegevica - Nomitel: Fired set of cattegsvis $5 GUN, Blood gHoup » coley osdival: which con have Rants. sj Best, eter , worse, coast - scale of Measuvernest of Data: ® Nowninal Scale data: # ava lifective rs catgorica| | # ° Gender, Labels. & okey doesn't matty: Red > 55 => 50% Blue > lL. 20% yellow > 3 2 30% ® oidinal Scale Detta :: + Ranking and ods matter. 4 viffeence qanot be wneasured: See Bate 5 3 hs cael Brech —> * woth > & @ Twiesvol scale dota: # nk and ovdynatters + Diferene con be mreasosed (ecloding wetio) # Dees not have “a” stayting vale - Si Tempeactuie- 30 F 6 F wo F fo F ® Ratio Scaled dota. - sods ond yonk matte. + Difference ¢ Ratio measurable . # am have “o” starting veloc. 9 MMS Yoo) F0, 65, 25, 90,42 measures of certtval Tendency *® Mean Mode Fe fedian —> Mean: The Average of the given jumbeds - x =§t, 2/2, 88) 9%] Meat = V4 DA DHBASA THO | = Wid Median + The ceritial element data of'te¥ sorting all the Values - Sst x2 {ry dort IF, Mateo} + Sort the values x={ Ie, 17,4, 2), 30, Wool Total ¢ Dato points ie xen) rombes * So. median ewil be mean of central Octo. Medion = E21 > bel x2{ 21,30, 85, 3, 10}, soxted > x= F083, 1G 30451 eX 5 odd noch: Data points. aso ynedian is centeY ekwerit, he s3° Mode: t's the most accusing Data pom m the given Dette. X= y ', a, 3,@ 2,2, 6, 6QO% 1BOQ am the above, (&| Appeared [6] Times So, Mode IS Measuies of Dispession Voos\arnee : Tt is the Average of Svrnmdlion of — Scyuares of differen belaeen each dato, point {o the mean of the Dela, points. Population Vvienee@") Samp le Vawiance (3) a) _ = Co ec a nel N where, aheye, xo: Dots, points. xp Data points a Flipping a com — Relling a dic Sets: Let's conside¥ 2 sets oe A=f 113,45, 6, 4,3,11 B=} 64,3), D Brtessection (AN8) : arts the common dota on Beth sets.A eB . fe %} D> a a = x vnion (au) : It’s combination of All dots Points ef A and 8B: AUB OB $5,2.3,45,.6 787} 3) subset : A is gubset of 8 —> false BIS subset of A —> True L) supesset: A is superset of 8 —> Twe gis spect of A —> False Covariance ! | TE is a aetyic te find how tao vasiables | axe elected to mach other: Hf cowxiane is iver: > They if XD | y 4t58=4 also if xl, y also L TH coveXionce Is -Ve: —>Thn, ff xv, y? if x7, Then yl | TP covasionce is 2eX0+ sit -vepresents thee fs no releation bla x andy. Advertage pisod varitage Finds Relectionship bo “doesn't have Limit X and Y Velves solos) EEDED yn-l roy ek PF OEY) x 3 -2 = 5 Gy uy Ss 8 ° 8 6 4 1 oy & x7y yes both aime Z + 3G>w Ais es ave, f xt,then y also 7 if xl, then yalso v Reason coxeleation: give US aneleation Rasnges bjs =| and! res ae tye, then thee Is te cexeheiticn rt ail a val then thee Is -Vve corleation . eH -Ve, & F ‘rem, then thee 1 Ne coreleation- _ cv) 56.9) aS speavnan coveleation: Tystead of xy oe alll we Rk), nk) & the peaxson axtlation fox this. A) RG) _ cav(RO), RG) pean Recital EAA J : \ SB) SR (x) RG) 3 4 FUN DY ms 2 ! we ye Histogsarn: a Tes te representation of the spreod of Decta Points - wwe have to da % symothening Cue te get the probability pistsibulion — cusve 3. let's foke the dela as xed Ty) 9, Wy Mey UL 2, 2S, 32,43, WS, 464747 SI, ssh lets Take bin size =5 OR. <5), (prabobil pistsibytion cusve) Types of Hislogc PDF cusves: Belocy axe some basic pistvibutions Noswna /@uassian + Iss called 95 symentsic A pists bution Skeaness : # Skegness (So vnetxic to undexstand the type of data distyibution- # oe can make hettes statistical decissions by understanding it: No skeaness: 1s : Here mean, wedian and Mode axe almast sini lox: x +Vve skecaness/ ight skeasess : yn Meat > Medion > Mode nde oy Neon, % -ve skeaness/Left y Skeamnes § - mode 7Median >Mea¥1 Probability Mass Function : * used to Represent pisctele Random vasiables Fos eg: Rolling Dice Xf 1°23) 48,6} pci) = V6 , PQ) = Ve , tae ft _ 4 Ts the combUve som of all the Data poirits. Probably ensty Fc + used to yepseseril continovs Random ae pistubutian- Bernoulli pistvibution « * Tt i the Distyibution aheve tado Discxete outcomes cuill be sepresenited by ep and |-P eg: Tossing 4 colyy p(t) =) P = 0°50 3 POs PP meat of Rexnoulli pistsibution + P Vorlance + Pay std deviation : JPy Binomial —Distei bation : * Tk iS Some aS pemoull’s But tao Discucte eutcomes aaill be done Fas [WW] nemnber of Times: gj, Toss Oo colt) 13. Times Neon : 0 Vosrig Nice : npy std Deviation > fnpar Poisson Distribution: # Ddescsibes the number of events causing at | a cextaity period of Time: eg: No of people visiting ospitad every [fe] hoor numbes ase Azik38S pee #(A) Represent S ai expected no: of eve! loo at every Time i tg see ig hime Tnitesval! Mean > At Vosience —> At # ts tine Trew Ae Expected no of: Everit’S Nesmal Gx) Gaussian pistsibation : gel eye g syrnetoc pstu bation ERE YT we A UE 3B ie Empisical Rule : 4 63% Data lies fo dn © and Ato % 957 Dela lies blo deze ond Ate a ae? dda lis blo arse ad Jub3e gs: Based on so much of AnalyS tS, Reseaxchess Found Height , ceights iis etc» ill mnestly have — guassian pistribution uniform oist¥ibution continous oniform distal but ton: & In uniform pistubution, the probability of getting cvtcoms Is ava # Tn continous unifesm pistsibution, In betacen a specified Range the probability f. getting outcomes is Same: Py= (x%.-%) eo b-a a= lowes bound b= highes bout - x oXq are dataponts : pu(x blo uo us) = (#A.- %) na “Oa 1 = (US We) FT = s/r0 = Wy a 028 328% (€3) SRM INSTITUTE OF SCIENCE & TECHNOLOGY @) a 1 a(x bles 30 @35) = (38 30) +e = O21 > psx] pr(x blo 20 us) > (8-30) #1_ Lo mea 3 OTA - 'S/r0 Voxxiomce * b-0)'/n 2 OFS LFS? piscaete onifosmn —vistwrbution: + All the data has cual probability te, Yn e9: Rolling a die { Von iV tee yn a a Se £31) 2 Yn 4 Ve standosd Newrnal oistebution : % Th is the Peocess of converting Nowa pists bution into Ve, 2% and +ve Dolo Fos ej this is Novwal pisty button : ' : a a) => d=o ae 6S} \ i ee Or ha pens 4S cg Let's apply 2 Scoxe on all Oxta w= mean > 3 = x7 ok es a = 2st§ Deviation ~2 aa data X- Scove © oie -I oS 0 oS : cag “EOS 88S ELS Ss NM SOW EY, X-ScoBeS » is used to standasd'ae the Data: F F-Scose boy eR oe + we can also yeleate “e) alth hoa mang std deviaction points a Data point 15 acy fyorn eat: ey issn wpind px of Region above 38 4 o-6tiS Yrts the veg lon out of W Befose 3-S abe 35% I-O-6HS 3 073085 2 30-387. > Region chi Savo%se + TE gs a method used te check hoor efficient axe 2 sets of data- He Co£) 4s E O= obsexved date. c= Expected dota. ss ~< expected ob&xved (© -e) 6-8 Ws “1S0 os s bs 200 S6IS «2. 4S nz 1S loo 6s S oe 50 ses uS ol = 0:05 (iven) 100 fey def 3) 42005, | ciitical value = #3 degite of frcedow > ne! =u! 20 ew chi suave abl, as/. 00S cet pp Reed ten +3! fis awe get chi sate Yale ae get s gate than #31 ie, loo > F8/, we ca covsidey evpected dato. is std ditfevent fom obsevel deta. F Test 3 we Tt is used to Ayd the compawision of variances of two dato set Fe : =(si) Se) 29 clos $ sie Vostance A 20 us 8 20 V3 O03 pejected Mea deqice of Pucedort\s : df, 2p Wale Lor! = l4 dfy = Helse Le- | 19 At dzo-08 and of, ="%, df,=!9, Pom F-Test fable, cxitica! value Is 2-20 ASF valve (1:39) ¢ citi Valve (2:0) we can say these is Mo significant difference. bla chss A and chss B.

You might also like