Kev Nkag Siab thiab Cov Ntsiab Lus Tseem Ceeb ntawm Kev Piav Qhia Txog Kev Txheeb Xyuas Cov Ntaub Ntawv
Cov ntaub ntawv piav qhia yog ib qho tseem ceeb tshaj plaws hauv cov txheej txheem tshuaj xyuas cov ntaub ntawv. Ua ntej ib tug neeg txiav txim siab, kwv yees, lossis txiav txim siab raws li cov ntaub ntawv, thawj kauj ruam yog yuav luag txhua zaus "nkag siab cov ntaub ntawv" nws tus kheej. Qhov no yog qhov uas cov ntaub ntawv piav qhia los ua si: pab sau ua ke, teeb tsa, thiab nthuav qhia cov ntaub ntawv kom nws cov qauv, cov yam ntxwv, thiab cov qauv tuaj yeem pom tseeb. Tsab xov xwm no tham txog kev txhais ntawm cov ntaub ntawv piav qhia thiab nws cov ntsiab lus yooj yim siv dav hauv kev tshuaj xyuas cov ntaub ntawv.
Kev Nkag Siab Txog Cov Txheeb Cais Piav Qhia
Feem ntau, kev txheeb cais piav qhia yog ib ceg ntawm kev txheeb cais uas tsom mus rau kev sau, kev sau luv luv, kev teeb tsa, thiab kev nthuav qhia cov ntaub ntawv los muab ib daim duab meej ntawm nws qhov xwm txheej. Nws lub hom phiaj tseem ceeb tsis yog los sim cov kev xav lossis ua kom dav dav rau cov pej xeem dav dua (qhov ntawd yog thaj chaw ntawm kev txheeb cais piav qhia), tab sis es piav qhia txog dab tsi tshwm sim hauv cov ntaub ntawv ntawm tes.
Piv txwv li, yog tias ib lub tsev kawm ntawv sau cov qhab nia xeem lej los ntawm 200 tus tub ntxhais kawm, cov ntaub ntawv piav qhia tuaj yeem siv los teb cov lus nug xws li: Qhov qhab nia nruab nrab yog dab tsi? Muaj kev sib txawv ntau npaum li cas hauv cov qhab nia? Cov qhab nia siab tshaj plaws thiab qis tshaj plaws yog dab tsi? Feem ntau cov qhab nia puas sib sau ua ke hauv qee qhov ntau thiab tsawg? Cov lus nug no tseem ceeb ua lub hauv paus rau kev ntsuam xyuas, yam tsis tas yuav kos cov lus xaus txog cov tub ntxhais kawm ntawm lwm lub tsev kawm ntawv.
Lub Luag Haujlwm ntawm Cov Txheeb Xyuas Piav Qhia hauv Kev Tshawb Fawb Cov Ntaub Ntawv
Hauv kev xyaum tshuaj xyuas cov ntaub ntawv, cov ntaub ntawv piav qhia feem ntau yog thawj kauj ruam uas txiav txim siab qhov kev taw qhia ntawm kev tshuaj xyuas tom qab. Nws lub luag haujlwm suav nrog:
1. Sau cov ntaub ntawv raw kom luv thiab yooj yim to taub.
2. Txheeb xyuas cov qauv xws li cov qauv sib txawv, cov pab pawg ntaub ntawv tseem ceeb, lossis cov teeb meem txawv txav.
3. Nrhiav cov ntaub ntawv yuam kev xws li cov nqi tsis tsim nyog, cov ntaub ntawv ploj lawm, lossis cov ntaub ntawv rov ua dua.
4. Nthuav qhia cov ntaub ntawv sib txuas lus los ntawm cov lus qhia, cov duab qhia, thiab cov ntsiab lus suav lej.
5. Txhawb kev txiav txim siab thaum ntxov, piv txwv li txiav txim siab cov tswv yim lag luam raws li cov ntaub ntawv ntawm cov neeg siv khoom.
Yog tsis muaj cov kauj ruam piav qhia, kev tshuaj xyuas ntxiv yuav tsis raug vim tias cov ntaub ntawv tsis tau nkag siab tag nrho.
Cov Hom Ntaub Ntawv thiab Cov Nplai Ntsuas
Lub tswv yim yooj yim ntawm kev piav qhia txog cov ntaub ntawv tsis tuaj yeem sib cais los ntawm kev nkag siab txog cov hom ntaub ntawv thiab cov nplai ntsuas, vim tias ob qho tib si txiav txim siab txoj kev sau ntawv kom raug.
1. Cov Ntaub Ntawv Zoo thiab Ntau
– Cov ntaub ntawv zoo (pawg): cov ntaub ntawv nyob rau hauv daim ntawv ntawm pawg lossis daim ntawv lo, piv txwv li poj niam txiv neej, txoj haujlwm, pawg khoom.
– Cov ntaub ntawv suav lej: cov ntaub ntawv ua cov lej uas suav tau lossis ntsuas tau, piv txwv li hnub nyoog, nyiaj tau los, qhov siab.
2. Kev Ntsuas Qhov Ntsuas
– Nominal: tsuas yog qhia txog pawg neeg xwb (piv txwv li: hom ntshav).
– Ordinal: muaj ib qho kev sib lawv liag, tab sis qhov deb ntawm cov pawg tsis paub meej (piv txwv li: qib kev txaus siab: qis-nruab nrab-siab).
- Lub sijhawm luv: qhov deb ntawm cov nqi yog tib yam, tab sis tsis muaj qhov xoom kiag li (piv txwv li: Celsius kub).
- Qhov sib piv: qhov deb yog tib yam thiab muaj qhov xoom kiag li (piv txwv li: qhov hnyav ntawm lub cev, cov nyiaj tau los).
Kev txiav txim siab qhov ntsuas ntawm cov ntaub ntawv yog qhov tseem ceeb rau kev xaiv cov kev ntsuas tsim nyog ntawm qhov sib txawv ntawm nruab nrab, kev ntsuas ntawm kev tawg, thiab kev pom.
Kev Nthuav Qhia Cov Ntaub Ntawv: Cov Lus Qhia thiab Cov Duab Qhia
Cov ntaub ntawv piav qhia feem ntau yog txuam nrog kev nthuav qhia cov ntaub ntawv kom nws yooj yim nyeem thiab txhais.
1. Rooj faib zaus
Ib daim ntawv qhia txog zaus uas qhia tias tus nqi lossis pawg twg tshwm sim ntau zaus. Qhov no pab tau rau cov ntaub ntawv loj, uas ua rau kom luv luv. Rau cov ntaub ntawv lej, zaus feem ntau raug teeb tsa ua ntu zus (piv txwv li, 0–10, 11–20, thiab lwm yam).
2. Cov Duab thiab Cov Duab Qhia
Qee hom kev pom kev uas pom tau ntau:
- Daim duab qhia kab: tsim nyog rau cov ntaub ntawv categorical.
- Daim duab qhia txog lub voj voog: qhia txog qhov sib piv ntawm txhua pawg (txawm hais tias rau ntau pawg nws feem ntau tsis zoo).
- Histogram: zoo ib yam li daim duab qhia kab tab sis rau cov ntaub ntawv lej uas tau muab faib ua pawg; pab kom pom cov duab ntawm kev faib tawm.
– Zaus polygon: ib txoj kab txuas cov zaus ntawm txhua chav kawm.
- Boxplot (daim duab thawv): qhia txog qhov nruab nrab, quartiles, kev faib tawm, thiab cov outliers muaj peev xwm.
Kev pom kev pab pom cov qauv lossis qhov tsis zoo ntawm cov ntaub ntawv uas qee zaum tsis meej yog tias koj tsuas yog saib cov lej xwb.
Kev Ntsuas ntawm Central Tendency
Kev ntsuas ntawm qhov sib txawv ntawm nruab nrab piav qhia txog tus nqi "nruab nrab" lossis tus nqi uas sawv cev rau cov ntaub ntawv zoo tshaj plaws.
1. Nruab Nrab (Nruab Nrab)
Qhov nruab nrab yog qhov sib sau ua ke ntawm txhua tus nqi faib los ntawm tus lej ntawm cov ntsiab lus ntaub ntawv. Qhov nruab nrab yog nrov vim nws yooj yim to taub, tab sis nws rhiab heev rau cov outliers. Hauv cov ntaub ntawv nyiaj tau los, piv txwv li, ib tus neeg nplua nuj heev tuaj yeem ua rau qhov nruab nrab tsis sib xws.
2. Nruab Nrab (Tus Nqi Nruab Nrab)
Tus lej nruab nrab yog tus nqi nruab nrab tom qab cov ntaub ntawv tau raug cais tawm. Yog tias tus lej ntawm cov ntsiab lus ntaub ntawv yog tus lej txawm, tus lej nruab nrab yog tus nqi nruab nrab ntawm ob tus nqi nruab nrab. Tus lej nruab nrab tiv taus cov outliers ntau dua, yog li nws feem ntau siv rau cov ntaub ntawv nrog kev faib tawm asymmetric.
3. Hom (Tus nqi tshwm sim ntau zaus)
Hom yog tus nqi uas tshwm sim ntau tshaj plaws thiab nws muaj txiaj ntsig zoo rau cov ntaub ntawv categorical. Piv txwv li, hom ntawm cov khoom uas yuav ntau tshaj plaws qhia txog qhov kev nyiam ua ntej.
Kev Ntsuas ntawm Kev Tawg Tawm
Ntxiv rau qhov paub txog tus nqi hauv nruab nrab, nws tseem ceeb heev kom paub tias cov ntaub ntawv kis tau los ntawm qhov chaw nruab nrab li cas.
1. Qhov ntau thiab tsawg
Qhov ntau thiab tsawg yog qhov sib txawv ntawm qhov siab tshaj plaws thiab qhov tsawg kawg nkaus. Qhov kev ntsuas no yooj yim, tab sis nws raug cuam tshuam los ntawm cov outliers.
2. Kev Hloov Pauv thiab Kev Hloov Pauv Tus Qauv
- Variance ntsuas qhov nruab nrab squared deviation ntawm cov nqi los ntawm qhov nruab nrab.
- Tus qauv sib txawv yog lub hauv paus plaub fab ntawm qhov sib txawv, feem ntau siv vim nws cov chav nyob zoo ib yam li cov ntaub ntawv thawj.
Qhov loj dua ntawm qhov kev hloov pauv tus qauv, qhov ntau dua ntawm cov ntaub ntawv hloov pauv; qhov me dua nws, qhov ntau dua ntawm cov ntaub ntawv feem ntau sib sau ua ke nyob ib puag ncig qhov nruab nrab.
3. Quartiles thiab IQR (Interquartile Range)
Cov Quartiles faib cov ntaub ntawv ua plaub ntu sib npaug zos:
- Q1 (qis dua quartile), Q2 (nruab nrab), Q3 (sab saud quartile).
IQR = Q3 − Q1 qhia txog kev faib tawm ntawm 50% nruab nrab ntawm cov ntaub ntawv, thiab nws tiv taus cov outliers.
Daim Ntawv Faib Tawm thiab Cov Outliers
Cov ntaub ntawv piav qhia kuj tseem yuav tsum tau them sai sai rau daim ntawv ntawm kev faib cov ntaub ntawv:
– Symmetrical: cov ntaub ntawv raug nthuav dav sib npaug zos rau sab laug thiab sab xis ntawm qhov nruab nrab / nruab nrab.
- Sab xis-skewed: ntau tus nqi me me, ob peb tus nqi loj.
- Sab laug-skewed: ntau tus nqi loj, ob peb tus nqi me.
Lub caij no, tus outlier yog tus nqi uas txawv ntawm feem ntau ntawm cov ntaub ntawv. Cov Outliers tuaj yeem tshwm sim vim yog qhov yuam kev sau lossis cov xwm txheej tseem ceeb hauv ntiaj teb tiag (piv txwv li, kev lag luam loj heev). Kev txheeb xyuas cov outliers yog qhov tseem ceeb vim tias lawv tuaj yeem cuam tshuam rau qhov nruab nrab, qhov sib txawv, thiab kev txhais lus tag nrho.
Xaus
Cov ntaub ntawv piav qhia yog thawj kauj ruam tseem ceeb hauv kev tshuaj xyuas cov ntaub ntawv vim nws pab hloov cov ntaub ntawv raw mus ua cov ntaub ntawv tseem ceeb. Los ntawm kev sau cov lej luv luv (qhov nruab nrab, nruab nrab, hom), kev ntsuas ntawm kev sib kis (ntau yam, kev hloov pauv tus qauv, IQR), thiab kev nthuav qhia cov ntaub ntawv hauv cov lus thiab cov duab, cov kws tshuaj xyuas tuaj yeem nkag siab sai thiab raug cov yam ntxwv ntawm cov ntaub ntawv. Kev nkag siab txog hom ntaub ntawv thiab qhov ntsuas ntsuas kuj txiav txim siab txoj kev piav qhia tsim nyog. Nrog lub hauv paus no, kev tshuaj xyuas tom qab - suav nrog kev tshuaj xyuas kev xav thiab kev txiav txim siab - tuaj yeem ua tiav hauv txoj kev tsom mus rau thiab lav ris ntau dua.
Yog tias koj xav tau, kuv tuaj yeem hloov kho tsab xov xwm no kom haum rau kev kawm ntau dua (nrog rau kev hais txog), yooj yim dua rau blog, lossis suav nrog cov piv txwv xam lej yooj yim thiab cov duab kos hauv daim ntawv teev/daim duab.
