Ingabe ungumhlaziyi wedatha ofuna ukuhlukana nesayensi yedatha? Uma kunjalo, lokhu okuthunyelwe kukuwe.
Abantu abaningi baqala kuma-analytics ngoba ngokuvamile banesithiyo esiphansi sokungena, kepha njengoba bethola isipiliyoni, bayabona ukuthi bafuna ukuthatha izinselelo eziningi zobuchwepheshe, noma vele bakhulise amandla abo okufunda. Ukuhamba kusuka kumhlaziyi wedatha kusosayensi wedatha kungaba ukuhamba komsebenzi ohlakaniphile – kepha kudinga isu elifanele.
Uma umusha lapha, igama lami nguMarina. Ngingusosayensi osetshenzisiwe e-Amazon, futhi ngisize inqwaba yabantu ukuguqukela kwezobuchwepheshe, ngisho nezizinda ezingezona zendabuko – zifakiwe.
Kulokhu okuthunyelwe, sizomboza konke okudingeka ukwazi ukwenza ukuguqulwa kusuka kumhlaziyi wedatha kusosayensi wedatha kuphumelele:
- Yimaphi amakhono ozodinga ukuwathuthukisa
- Izinsizakusebenza Zami Ezithandekayo Zokufunda
- Kanye namasu okuxoxisana okufika nokuthola okunikezwayo kwemisebenzi
Ake singene kukho, siqale ngokunquma ukuthi ngabe lolu shintsho luwumqondo omuhle yini kuwe kwasekuqaleni.
Ukuqhathanisa indima
Ngaphambi kokuthi siqale, ake siqinisekise ukuthi sonke sisekhasini elifanayo mayelana nokuthi umehluko phakathi kwalezi zindima noma kunjalo, uqala nge-data analytics.
Abahlaziyi bedatha bagxila ekusebenzeni nedatha ehlelekile ukushayela izinqumo zebhizinisi. I-Toolkit yabo ifaka i-SQL, i-Excel, iTableau noma i-Powerbi, kanye ne-Python eyisisekelo yokucubungula idatha, ukubona ngeso lengqondo, futhi mhlawumbe nokuhlaziywa kwezibalo okulula. Izindima ezigxile ekuqondeni ngokomlando Idatha yokuphendula imibuzo ngokwenzekile nokuthi kungani.
Ososayensi bemininingwane bakha kulezi zisekelo kodwa banwebeka babe yimodeli yokuqagela kanye nokwenza izinqumo ezenzakalelayo. Ngenkathi futhi zisebenzisa i-SQL nePython, zisebenza kakhulu ngokutholwa kwezibalo, izinhlaka zokufunda ngomshini, namapulatifomu efu. Ukugxila kwabo kuguqukela ekubikezeleni -phambili Imiphumela nokuncoma izenzo.
Umbono ongemuhle ojwayelekile wokuthi abahlaziyi bedatha kumele babe ososayensi bedatha ukuqhubekisela phambili imisebenzi yabo. Impela lokho akulona iqiniso!
Abahlaziyi abakhulu bangathola amaholo aphezulu futhi banomthelela webhizinisi oqinile ngaphandle kwe-ML ejulile noma ulwazi lwezibalo.
Ngokweqiniso, akuwona wonke umuntu ozojabulela umsebenzi wesayensi yedatha, futhi abaningi bangajabula ukuhlala endleleni yokuhlaziya.
Ngakho-ke ngaphambi kokuthi siqhubeke, zibuze le mibuzo elandelayo:
- Ingabe ufuna ukwazi ngokufunda komshini nokuthi isebenza kanjani?
- Ingabe ukhululekile nge (noma okungenani uthande) izibalo ezithuthukile kanye nezibalo?
- Ingabe ukhululekile ngezinselelo zezobuchwepheshe kanye nemiqondo yobunjiniyela yesoftware?
- Ngabe ulungile ngeqhaza elinobuqili obuningi, emsebenzini wansuku zonke kanye nokuthuthuka komsebenzi?
Uma usunami futhi ucabanga ukuthi “Yebo, ngifisa ukuphishekela isayensi yedatha,” Masikhulume ngokuthi senzeke kanjani ngempela.
Amakhono adingekayo ekuguqulweni
Kulungile, manje njengoba usunqume ukwenza ushintsho, ake sibhidlize amakhono asemqoka uzodinga ukuthuthukisa. Sizogxila ezindaweni ezine ezisemqoka ezakha isisekelo somsebenzi wesayensi yedatha.
Mathematics & Izibalo
Uma uvela ngemuva kwe-analytics ingemuva, kungenzeka ukuthi uvezwa okuthile kwezibalo, kepha isayensi yedatha ingadinga ukujula okuthe xaxa kwi-Math Front. Uzodinga ukukhululeka nge:
- Calculator Calculatorable and Linear Algebra, ikakhulukazi imisebenzi ye-matrix kanye ne-gradients yokuqonda ama-algorithms wokufunda umshini. Kepha ungakhathazeki – awudingi ukuba uchwepheshe wezibalo, udinga nje ngokwanele ukuqonda izisekelo zokukusiza uqonde ukuthi ama-algorithms asebenza kanjani.
- Uzodinga futhi i-Theory theory ne-Hypothesis Test yokuhlolwa kokuhlola.
- Kanye nemiqondo yezibalo njengezinhlobo ezahlukene zokusatshalaliswa kanye namasu wokuphindaphinda
- Futhi okuhle, okunye okuhlangenwe nakho ngokutholwa kokulungisa
Ukuhlahlela
Uma usuvele usebenzisa i-SQL ne-Python eyisisekelo endimeni yakho, uqala ikhanda lapha. Manje kumayelana nokukhuphuka. Gxila ku:
- I-Python ethuthukile kakhulu, ngakho-ke izinto ezifana nezisekelo ze-OOP, ukubhala ikhodi yokulondolozwa kwe-modular, ukuhlolwa kweyunithi, ukusebenza kahle kokusebenza, nokunye.
- Kusetshenziswa uhlaka lwe-ML njenge-skikit-funda, i-tensorflow, ne-pytorch.
- Futhi ukujwayelana nezinhlaka zedatha eziyisisekelo nama-algorithms wezingxoxo zamakhodi. Ngokuvamile lokhu kuzoba yimibuzo nje e-art
Umshini wokufunda kanye nezisekelo ze-AI
Le enye insika ebaluleke kakhulu yesayensi yedatha, ngakho-ke uzofuna ukukhululeka nge-ML Okuyisisekelo esifana:
- Ukufunda okugadiwe (Ngakho-ke, regression kanye nokuhlukaniswa).
- Ukufundwa okungahleliwe (izinto ezinjengokuncishiswa kokuhlangana nokuncipha).
- Ukuhlolwa kwemodeli kanye nokuqinisekiswa.
- Izisekelo zokufunda ezijulile.
- Futhi kulezi zinsuku, ukujwayela iGenai kuyinto futhi (kodwa ngalokhu kusho ukufunda ukufunda nama-API, hhayi amamodeli wokuqeqesha kusuka ekuqaleni)
Idatha enkulu nemiqondo yobunjiniyela bedatha
Ekugcineni, izindima eziningi zesayensi yedatha zibandakanya ukusebenza ngemininingwane emikhulu namapayipi ezenzakalelayo. Kulokhu, uzofuna ukugxila ku:
- Ukusebenza ngamapulatifomu e-computer wefu, ikakhulukazi izinsizakalo ze-AWS ezinjenge-S3 ne-sagemaker
- Ukuthuthukiswa kwe-Pipeline Development Kusetshenziswa amathuluzi anjenge-airflow
- Imigomo yokuqamba yohlelo oluyisisekelo lokulinganisa izixazululo zakho (lokhu kubaluleke kakhulu njengoba uphezulu kakhulu noma ugxile kakhulu ku-ML).
Ungawathuthukisa kanjani la makhono
Manje njengoba sesimboziwe ini Udinga ukufunda, ake sixoxe ngokuthi ungazakha kanjani la makhono. Kunezindlela ezimbalwa ezahlukahlukene ongazithatha, futhi okulungile kuwe kuzoncika kwisabelomali sakho, isitayela sokufunda, kanye nohlelo.
Ukuzifundela
Uma uzikhuthazela futhi uqondiswe kabusha, ukuzitadisha, ukuzitadisha kungaba yindlela efinyelelekayo futhi engabizi kakhulu yokuguqukela kwisayensi yedatha. Isihluthulelo somkhuba ongaguquki futhi ukhetha izinsizakusebenza ezifanele.
Nazi ezinye izifundo ezinhle engizozincoma ukuphuma, ngokulandelana (lezi yizixhumanisi ezihambisanayo, BTW!):
Uzodinga futhi ukuthola ukuqonda kwe-DSA eyisisekelo yokuthola amakhodi we-prep. Kulokhu ngikujabulele ukufundisa okufundisayo amaphethini wokuxoxisana we-coding ePython, ogxile emaphethini ajwayelekile wezakhiwo zedatha nemibuzo ye-algorithms. Ngithole lokhu kusiza kakhulu ukuze kubonakale sengathi udinga ukudinga ukuthi “wazi iqhinga” ukuphendula inkinga ye-leetcode.
Futhi, izincwadi ezimbalwa okufanele zifunde (lezi zixhumanisi ezihambisanayo, kepha ngi- <3 zonke lezi zincwadi):
Kunamathani amaningi, kepha lokhu kungaba ezintathu zami eziphezulu. Nasi isixhumanisi sezincwadi eziningi ezithandwayo zobuchwepheshe uma ufuna ukuhlola ngokwengeziwe!
Into ebaluleke kakhulu lapho uhamba indlela yokuzifundela ukuzihlola kungukuvumelana. Yenza uhlelo bese unamathela kulo, noma ngabe kuncane nje usuku ngalunye.
Amabhulurcamp
Manje, mhlawumbe ucabanga ukuthi ungathanda ukuba nesakhiwo esithe xaxa nokuziphendulela kwangaphandle ekufundeni kwakho. Uma ungafuni ukuzibophezela ngezinga eligcwele, ama-bootcamp angaba enye inketho.
Ezinye izinto ze-bootcamp yilezi:
- Ukufunda okusheshayo kwe-paced – ngokuvamile ungaziqedela ezinyangeni ezimbalwa.
- Ikharikhulamu ehlelekile, ngoba konke kubekelwe wena, ngakho-ke akudingeki ukuba uqede ndawonye uhlelo lwakho lokufunda.
- Kanye nokusekelwa komphakathi – uthola ukufundela kontanga futhi uthole ukuqeqeshwa kwabafundisi okungenzeka ukuthi abantu base besebenza endle.
Into eyodwa okufanele uyigcine engqondweni ukuthi ama-bootcamp ahluka ngekhwalithi, futhi akubona bonke ababaluleke kakhulu ngabaqashi. Ngaphambi kokubhalisa, yenza ucwaningo lwakho – ngakho-ke, hlola ukubuyekezwa, khuluma ne-alumni, futhi uqiniseke ukuthi anikela ngokusekelwa komsebenzi.
Isiqu esiphezulu
Kulabo abafuna ukungena okujulile kwisayensi yedatha ngamathuba aqinile okuxhumana, iziqu ze-master zingaba ukutshala okuqinile. Lokhu kuwusizo ikakhulukazi uma uguqulwa kusuka kwisizinda esingewona sezobuchwepheshe, noma uma ukhathazekile ingemuva lakho ngeke kube kudluliswa amathuluzi okuskena.
Okubi kakhulu kusobala ukuthi izinhlelo zikaMbusi zingabiza futhi zidle isikhathi. Kepha izindaba ezinhle lapha ukuthi manje sekunezinhlelo ezishibhile, zesikhathi se-inthanethi esikuvumela ukuthi utadishe ngenkathi usebenza. Isibonelo, izinhlelo zikaGeorgia Tech zibizeleka ngempela futhi zisezingeni elifanele.
Ukufundisa
Akunandaba noma iyiphi indlela oyithathayo, ukuqeqeshwa kungasiza ngendlela emangalisayo. Ukuba nomuntu okufanele ukukuqondise, ahlinzeke ngempendulo, futhi asize ngokuhamba ngomsebenzi kungenza umehluko omkhulu.
Ezinye izindlela zokuthola abeluleki:
- Ngenkampani yakho – Uma inkampani yakho inososayensi wedatha, buza ukuthi ungabambisana yini noma ubathumele yini.
- I-LinkedIn – Joyina amaqembu weSayensi yedatha noma afinyelele ochwepheshe (nginevidiyo yonke kumasu okufundisa uma udinga usizo ngalokhu!).
- Imiphakathi eku-inthanethi efana ne-reddit, amaseva we-discord, namaqembu ase-slack angaba enye indlela yokuxhuma nabanye abafundi kanye nochwepheshe.
- Noma, qasha umeluleki – uma uthatha izinto ngokusekela ngokushesha, ukutshala imali kumeluleki -ithini kufanelekile.
Ukubonisa isipiliyoni
Kulungile, ngakho-ke ufunde wonke amakhono owadingayo. Kuhle lokho, kepha ufakazela kanjani umqashi ongaba ngumqashi ongenzayo empeleni umsebenzi wososayensi wedatha?
Nginevidiyo ephelele yokuthi ungakha kanjani iphothifoliyo futhi ngithole isipiliyoni ngaphandle komsebenzi wakho wesikhathi esigcwele. I-TL; u-Dr kukhona ukuthi kufanele uzame okusemandleni akho ukwenza amaphrojekthi akhuthazayo akuvumela ukuthi ulingise izimo zokusebenza ngokuba semsebenzini eduze ngangokunokwenzeka.
Kepha uma ufunda lokhu okuthunyelwe, kunethuba elihle njengamanje osebenza ngalo njengomhlaziyi wedatha vele, okukunikeza elinye elinye lisethi yamathuba okuthola indima yakho.
Isibonelo, ake sithi udala imibiko njalo kwi-Excel noma iTandeau. Ungasebenzisa le nqubo ngemibhalo yePython, mhlawumbe noma engeza ezinye izinto zokubikezela. Noma uma inkampani yakho isebenza ngovivinyo lwe-A / B, ivolontiya ukusiza ngokuhlaziywa kwezibalo.
Uma uneqembu lesayensi yedatha, zama ukusebenzisana nabo kwiphrojekthi. Futhi uma kungekho iqembu lesayensi yedatha, ligibela umqashi wakho kumaphrojekthi athile anomthelela ongakusiza nokuthi ufunde.
Isimo esihle kakhulu, lokhu kungaholela ekushintsheni kwangaphakathi. Icala elibi kakhulu, manje usunezibonelo ezinhlelweni zekhonkrithi zethonya namaphrojekthi wesayensi yangempela yokufaka ekuqhubekeni kwakho kabusha.
Ukuthola Umsebenzi
Uma ukwazi ukuguqukela ngaphakathi ke kuhle, usuqedile! Uma kungenjalo, nazi amanye amasu okukusiza ukuthi uthole le ndima yesayensi yokuqala yedatha:
Okokuqala, ake sixoxe ngokuthi uzibeka kanjani online. Ukuqalisa kwakho kabusha, i-LinkedIn, futhi i-GitHub isidingo sokutshela izindaba ezingaguquki osuvele usuvele usosayensi wedatha (ngoba uma unamakhono futhi wenze amaphrojekthi aqinile, uwenzile!). Ngakho-ke, esikhundleni sokubhala “umhlaziyi wedatha efuna iqhaza ledatha yesayensi,” ungahle uthi “uchwepheshe wedatha ochwepheshe ngokukhethekile kokuhlaziya kokuqagela kanye nokufunda komshini.”
Uma kukhulunywa nge-github yakho, qiniseka ukubeka izinto zakho ezinhle phezulu lapha. Lokhu kubaluleke kakhulu kubahlaziyi, ngoba amakhono akho okufaka amakhodi azohlolisiswa ngokwengeziwe. Ngakho-ke,
- Phina amaphrojekthi wakho we-ML amahle kakhulu phezulu
- Bhala ukufundwa okucacile okuchaza indlela yakho
- Qiniseka ukuthi ikhodi yakho ihlelwe kahle futhi ibhalwe phansi, ikukhombisa ukuthi uqonde izimiso zobunjiniyela zesoftware
- Futhi engeza ukubona okubonakalayo kanye nemiphumela ukukhombisa umthelela, okufanele kube lula kuwe ngemvelaphi yakho!
Uma sekuyisikhathi sokusebenzisa, ukubeka phambili izindima ze-hybrid. Lezi yizikhundla ezihlala phakathi kokuhlaziywa kwendabuko kanye nesayensi yedatha, futhi zivame ukuba yitshe elihle kakhulu.
Isibonelo, izinkampani eziningi (kufaka phakathi amafemu amakhulu wezobuchwepheshe afana ne-meta ne-Amazon) zinezindima ezibize ngazo “isayensi yedatha” kodwa empeleni zifana nezikhundla zokuhlaziya ezithuthukile. Futhi ngokweqiniso ezinkampanini eziningi, imigqa iyimfihlakalo noma kunjalo. Sebenzisa leli mpiguity ukuze inzuzo yakho!
Lapho uxhumana nenethiwekhi futhi ulungiselela izingxoxo, thola isizinda sakho se-analytics. Sebenzisa ukuqonda kwakho okujulile kwesimo sebhizinisi, amakhono okuxhumana acacile, kanye nezibonelo zendlela othonye ngayo ibhizinisi ukuletha umthelela ongalingani. Abanye abavoti okungenzeka babe ngezobuchwepheshe ukwedlula indlela ongaba khona ngebhizinisi nezokuxhumana kwezinto. Ngakho-ke ungesabi ukuncika ngamandla akho.
Khumbula, lolu shintsho ngeke luyenze ubusuku bonke, futhi kulungile. Okubalulekile inqubekela phambili engaguquki. Yonke imigqa yekhodi oyibhalayo, yonke imiqondo oyifundayo, yonke iphrojekthi oyigcwalisayo – konke kuyangezela.
Uma uzizwa sengathi udinga ukwesekwa okuthile nge-Data Science / ML Career, nazi izindlela engingasiza ngazo:
QAPHELA: Lokhu okuthunyelwe kuqukethe izixhumanisi ezihambelana nazo. Uma wenza ukuthenga ngizothola ikhomishini encane, kungabikho kubiza. Ngiyabonga ngokusekelwa kwakho ❤
