Isiqulatho[Fihla][Bonisa]
Ngendlela, sonke siyayazi ukuba iteknoloji yokufunda koomatshini iphuhliswe ngokukhawuleza kangakanani kwiminyaka edlulileyo. Ukufunda ngoomatshini luqeqesho olutsale umdla weenkampani ezininzi, izifundiswa kunye namacandelo.
Ngenxa yoku, ndiza kuxoxa ngezona ncwadi zibalaseleyo zokufunda koomatshini ekufuneka injineli okanye umntu oqalayo ukufunda namhlanje. Kufuneka ukuba nivumile nonke ukuba ukufunda iincwadi akufani nokusebenzisa ingqondo.
Ukufunda iincwadi kunceda iingqondo zethu zifumane izinto ezininzi ezintsha. Ukufunda kukufunda, emva koko. Ithegi yomfundi ozifundelayo kumnandi kakhulu ukuba nayo. Ezona ncwadi zesikhokelo zibalaseleyo ezikhoyo entsimini ziya kubalaseliswa kweli nqaku.
Ezi ncwadi zilandelayo zibonelela ngentshayelelo ezanyiweyo neyinyani kwinkalo enkulu ye-AI kwaye zihlala zisetyenziswa kwizifundo zaseyunivesithi kwaye zicetyiswa ngabafundi kunye neenjineli ngokufanayo.
Nokuba unetoni ye yokufunda umatshini Amava, ukuchola enye yezi ncwadi zesikhokelo kunokuba yindlela eyoyikisayo yokuxubha. Ngapha koko, ukufunda yinkqubo eqhubekayo.
1. Ukufunda ngoomatshini kubaQalayo ngokupheleleyo
Ungathanda ukufunda ukufunda ngoomatshini kodwa awazi ukuba kwenziwa njani. Kukho iikhonsepthi ezininzi ezibalulekileyo zethiyori kunye nezibalo ekufuneka uziqonde ngaphambi kokuba uqalise uhambo lwakho lwe-epic lokufunda ngomatshini. Yaye le ncwadi iyanelisa loo ntswelo!
Inika ii-novices ezipheleleyo ezinomgangatho ophezulu, osebenzayo intshayelelo yokufunda koomatshini. Incwadi yokuFunda ngoMatshini yabaQalayo ngokugqibeleleyo lolona khetho lubalaseleyo kuye nabani na okhangela eyona ngcaciso ilula yokufunda koomatshini kunye nezimvo ezinxulumeneyo.
Ii-algorithms zeml ezininzi zencwadi zikhatshwa ziingcaciso ezimfutshane kunye nemizekelo yegraphic ukunceda abafundi ukuba baqonde yonke into ekuxoxwa ngayo.
Imixholo equkunjelwe kule ncwadi
- Iziseko ze amanethiwekhi
- Uhlalutyo lohlalutyo
- Ubunjineli obubonakalayo
- Clustering
- Ukuqinisekiswa okunqamlezileyo
- Iindlela zokucoca idatha
- Imithi yezigqibo
- Imodeli ehlanganisiweyo
2. Ukufunda ngoomatshini kwiiDummies
Ukufunda ngoomatshini kunokuba ngumbono obhidayo kubantu abaqhelekileyo. Nangona kunjalo, ixabiseke kakhulu kuthi thina banolwazi.
Ngaphandle kwe-ML, kunzima ukulawula imiba efana neziphumo zophendlo lwe-intanethi, iintengiso zexesha lokwenyani kumaphepha ewebhu, i-automation, okanye ukuhluzwa kwe-spam (Ewe!).
Ngenxa yoko, le ncwadi ikunika intshayelelo ethe ngqo eya kukunceda ufunde ngakumbi malunga nommandla ontsonkothileyo wokufunda koomatshini. Ngoncedo lwe-Machine Learning For Dummies, uya kufunda indlela "yokuthetha" iilwimi ezifana nePython kunye ne-R, eya kukuvumela ukuba uqeqeshe iikhomputha ukwenza ukuqaphela ipateni kunye nohlalutyo lwedatha.
Ukongeza, uya kufunda ukusebenzisa i-Python's Anaconda kunye ne-R Studio ukuphuhlisa kwi-R.
Imixholo equkunjelwe kule ncwadi
- Ukulungiswa kwedatha
- iindlela zokufunda koomatshini
- Umjikelo wokufunda koomatshini
- Ukufunda okubekelwe iliso nokungajongwanga
- Iinkqubo zokufunda koomatshini boqeqesho
- Ukubopha iindlela zokufunda koomatshini kwiziphumo
3. Incwadi yokuFunda ngoMatshini alikhulu
Ngaba kunokwenzeka ukugubungela yonke imiba yokufunda koomatshini phantsi kwamaphepha ali-100? UAndriy Burkov's Incwadi yokuFunda yoMatshini wePhepha eliKhulu lilinge lokwenza okufanayo.
Incwadi yokufunda ngomatshini ibhalwe kakuhle kwaye ixhaswa ngabaphathi beengcamango ezidumileyo kuquka uSujeet Varakhedi, iNtloko yezobuNjineli kwi-eBay, kunye noPeter Norvig, uMlawuli woPhando kwiGoogle.
Yeyona ncwadi inkulu kumntu oqalayo ukufunda ngoomatshini. Emva kokufunda ngokucokisekileyo le ncwadi, uya kuba nakho ukwakha nokuqonda iinkqubo ze-AI eziphucukileyo, uphumelele kudliwanondlebe lokufunda ngomatshini, kwaye uqalise nenkampani yakho esekwe kwiML.
Nangona kunjalo, le ncwadi ayenzelwanga abaqalayo abagqibeleleyo ekufundeni koomatshini. Jonga kwenye indawo ukuba ufuna enye into ebaluleke ngakumbi.
Imixholo equkunjelwe kule ncwadi
- IAnatomy ka ukufunda algorithm
- Ukufunda okubekelwe iliso kunye nokufunda okungajongwanga
- Ukomeleza ukuFunda
- Ii-algorithms ezisisiseko zokuFunda koomatshini
- Isishwankathelo sothungelwano lweNeural kunye nokufunda okunzulu
4. Ukuqonda ukuFunda koomatshini
Intshayelelo ecwangcisiweyo yokufunda koomatshini inikwe kwincwadi ethi Ukuqonda ukuFunda koMatshini. Le ncwadi iphonononga nzulu kwiimbono ezisisiseko, iiparadigms zokubala, kunye nokuphuma kwimathematika yokufunda koomatshini.
Uluhlu olubanzi lwezifundo zoomatshini zivezwa ngendlela elula ngokufunda koomatshini. Iziseko zethiyori zokufunda koomatshini zichazwe kwincwadi, kunye nezinto eziphuma kwimathematika ezijika ezi ziseko zibe zii-algorithms eziluncedo.
Le ncwadi inikezela ngeziseko ezingundoqo ngaphambi kokugubungela uluhlu olubanzi lwezifundo ezibalulekileyo ezingakhange zigutyungelwe kwiincwadi zezifundo zangaphambili.
Okubandakanywe koku yingxoxo yeconvexity kunye nozinzo iikhonsepthi kunye nobunzima bokubala kokufunda, kunye ne-algorithmic paradigms ebalulekileyo njengestochastic. ukuhla komgangatho, uthungelwano lwe-neural, kunye nokufunda okucwangcisiweyo okuphumayo, kunye neengcinga ezisandula ukuvela zethiyori ezifana nendlela yePAC-Bayes kunye nemida esekwe kuxinzelelo. yenzelwe ukuqalisa amabanga okanye amabanga aphakamileyo.
Imixholo equkunjelwe kule ncwadi
- Ukuntsonkotha kokufunda koomatshini
- ML algorithms
- Unxibelelwano lweeNeural
- Indlela yePAC-Bayes
- Ukwehla kwe-Stochastic gradient
- Ukufundwa kweziphumo ezicwangcisiweyo
5. Intshayelelo yokuFunda koomatshini ngePython
Ngaba uyisazi sedatha yePython-savvy ofuna ukufunda ukufundwa koomatshini? Eyona ncwadi ilungileyo yokuqalisa uhambo lwakho lokufunda ngoomatshini sintshayelelo kwiSifundo soomatshini ngePython: Isikhokelo seeNzululwazi zeDatha.
Ngoncedo lwencwadi ethi Intshayelelo yokuFunda koMatshini ngePython: Isikhokelo seeNzululwazi zeDatha, uya kufumana iindlela ezahlukeneyo eziluncedo zokudala iinkqubo zokufunda zoomatshini.
Uya kugubungela lonke inyathelo elibalulekileyo elibandakanyekayo ekusebenziseni iPython kunye nephakheji yeScikit-Learn ukwakha usetyenziso oluthembekileyo lokufunda koomatshini.
Ukufumana ukuqonda okuqinileyo kwe-matplotlib kunye namathala eencwadi eNumPy kuya kwenza ukufunda kube lula kakhulu.
Imixholo equkunjelwe kule ncwadi
- Ubuchwephesha banamhlanje bokwenza iparameter tweaking kunye novavanyo lwemodeli
- Izicelo kunye neengcamango ezisisiseko zokufunda koomatshini
- iindlela zokufunda ezizenzekelayo
- Ubuchwephesha bokukhohlisa idatha yombhalo
- Imodeli yekhonkco kunye nemibhobho ye-encapsulation yokuhamba komsebenzi
- Ukumelwa kwedatha emva kokucubungula
6. Ukufunda ngoomatshini ngezandla nge-Sci-kit funda, iiKeras & Tensorflow
Phakathi kwezona zipapasho zicokisekileyo kwisayensi yedatha kunye nokufunda koomatshini, igcwele ulwazi. Kucetyiswa ukuba iingcali kunye nabaqalayo ngokufanayo bafunde ngakumbi ngalo mbandela.
Nangona le ncwadi iqulethe nje ithiyori encinane, ixhaswa yimizekelo eyomeleleyo, inika indawo kuluhlu.
Le ncwadi ibandakanya izihloko ezahlukeneyo, kubandakanywa i-scikit-learn yeeprojekthi zokufunda ngomatshini kunye ne-TensorFlow yokudala kunye nokuqeqesha i-neural networks.
Emva kokufunda le ncwadi, sicinga ukuba uya kukulungela ngakumbi ukuphonononga ngakumbi ukufunda okunzulu kwaye ujongane neengxaki eziphathekayo.
Imixholo equkunjelwe kule ncwadi
- Jonga umhlaba wokufunda koomatshini, ngakumbi uthungelwano lwe-neural
- Landela umkhondo weprojekthi yokufunda ngomatshini ukusuka ekuqaleni ukuya esiphelweni usebenzisa iScikit-Learn.
- Vavanya iimodeli zoqeqesho ezininzi, ezinje ngobuchule bokuhlanganisana, amahlathi angacwangciswanga, imithi yezigqibo, kunye noomatshini bevector yenkxaso.
- Yenza kwaye uqeqeshe uthungelwano lwe-neural ngokusebenzisa ithala leencwadi leTensorFlow.
- Qwalasela uthungelwano lwe-convolution, iinethi eziqhubekayo, kunye nokufunda okunzulu okomeleza ngelixa uphonononga umnatha we-neural ziyilo.
- Funda indlela yokukala kunye nokuqeqesha ubunzulu bothungelwano lwe-neural.
7. Ukufunda ngoomatshini kubaHackers
Kumdwelisi onamava onomdla kuhlalutyo lwedatha, kubhaliwe incwadi ethi Machine Learning for Hackers. IiHacker ziingcali zezibalo kulo mongo.
Kumntu onokuqonda okuqinileyo kwe-R, le ncwadi lukhetho olukhulu kuba uninzi lwayo lugxile kuhlalutyo lwedatha kwi-R. Ukongezelela kubandakanywa kwincwadi yindlela yokuguqula idatha usebenzisa i-R ephezulu.
Ukubandakanywa kwamabali achaphazelekayo kugxininisa ixabiso lokusebenzisa i-algorithms yokufunda koomatshini kunokuba yincwadi ethi Machine Learning for Hackers eyona ndawo ibalulekileyo yokuthengisa.
Incwadi inika imizekelo emininzi yehlabathi lokwenyani ukwenza umatshini wokufunda ube lula kwaye ukhawuleze kunokuba ungene nzulu kwithiyori yayo yemathematika.
Imixholo equkunjelwe kule ncwadi
- Yenza umhleli ongenalwazi weBayesian ohlalutya ngokulula umxholo we-imeyile ukufumanisa ukuba ngaba ngugaxekile.
- Ukuqikelela inani leembono zephepha kwiiwebhusayithi eziphezulu ze-1,000 usebenzisa ukuhlehliswa komgca
- Phanda iindlela zokwandisa ngokuzama ukukrazula unobumba othe ngqo we-cipher.
8. Python Machine Learning ngeMizekelo
Le ncwadi, ekunceda ukuba uqonde kwaye udale ukuFunda koMatshini ohlukeneyo, ukuFunda ngokuNzulu, kunye neendlela zoHlalutyo lweDatha, kusenokwenzeka ukuba yiyo kuphela egxile kwiPython njengolwimi lwenkqubo.
Iquka amathala eencwadi anamandla okuphumeza iialgorithms ezahlukeneyo zokuFunda ngoomatshini, njengeScikit-Learn. Imodyuli yeTensor Flow isetyenziselwa ukufundisa ngokufunda nzulu.
Ekugqibeleni, ibonisa amathuba amaninzi okuhlalutya idatha anokuphunyezwa ngokusebenzisa umatshini kunye nokufunda okunzulu.
Ikwakufundisa iindlela ezininzi ezinokuthi zisetyenziswe ukonyusa ukusebenza kwemodeli oyenzayo.
Imixholo equkunjelwe kule ncwadi
- I-Python yokufunda kunye nokuFunda koMatshini: Isikhokelo sabaQalayo
- Ukuphonononga isethi yedatha yamaqela eendaba ezi-2 kunye nokufunyanwa kwe-imeyile ye-spam ye-Naive Bayes
- Ukusebenzisa ii-SVMs, hlela izihloko zamabali eendaba Cofa-ngokuxela kwangaphambili usebenzisa i-algorithms esekwe kwimithi
- Uqikelelo lwesantya sokucofa usebenzisa uhlengahlengiso
- Ukusetyenziswa kwee-algorithms zokuhlehla ukuqikelela amaxabiso esitokhwe aphezulu ngemigangatho
9. Ukufunda koomatshini bePython
Incwadi yePython Machine Learning ichaza iziseko zokufunda koomatshini kunye nokubaluleka kwayo kwi-domain yedijithali. Yincwadi yokufunda ngomatshini yabaqalayo.
Okongeziweyo kugutyungelwe kule ncwadi kukufunda ngomatshini iindawo ezininzi ezingaphantsi kunye nokusetyenziswa. Imigaqo yenkqubo yePython kunye nendlela yokuqalisa ngolwimi lwenkqubo yasimahla kunye nomthombo ovulekileyo nayo igutyungelwe kwincwadi yePython Machine Learning.
Emva kokugqiba incwadi yokufunda koomatshini, uya kuba nakho ukuseka ngokusebenzayo inani lemisebenzi yokufunda koomatshini usebenzisa ikhowudi yePython.
Imixholo equkunjelwe kule ncwadi
- Izinto ezisisiseko zobukrelekrele bokwenziwa
- umthi wesigqibo
- Ukuhlengahlengiswa kwezinto
- Uthungelwano lwe-neural olunzulu
- Iziseko zolwimi lwePython
10. Ukufunda ngoomatshini: Imbono enokwenzeka
Ukufunda ngoomatshini: iProbabilistic Perspective yincwadi yokufunda ngomatshini ehlekisayo ebonisa imizobo yemibala ye-nostalgic kunye nokusebenza, imizekelo yehlabathi yokwenyani evela kumacandelo afana nebhayoloji, umbono wekhompyuter, iirobhothi, kunye nokucubungula umbhalo.
Igcwele iprozi eqhelekileyo kunye nepseudocode yee-algorithms ezibalulekileyo. Ukufunda ngoomatshini: Umbono onokwenzeka, ngokungafaniyo nolunye upapasho lokufunda koomatshini oluvezwa ngendlela yencwadi yokupheka kunye nokuchaza iindlela ezahlukeneyo ze-heuristic, kugxininise kwindlela esekelwe kumgaqo-siseko.
Ikhankanya iimodeli ze-ml kusetyenziswa umboniso wegraphical ngendlela ecacileyo neqondakalayo. Ngokusekwe kwindlela emanyeneyo, enokwenzeka, le ncwadi yesikhokelo ibonelela ngentshayelelo epheleleyo nezimeleyo kwindawo yokufunda koomatshini.
Umxholo ubanzi kwaye unzulu, kuquka imathiriyeli eyimvelaphi esisiseko kwizihloko ezinje ngokuba nokwenzeka, ukwenziwa ngcono, kunye nealjebra yomgca, kunye nengxoxo yenkqubela phambili yangoku kwindawo efana nemimandla engakhethiyo enemiqathango, uhlengahlengiso lwe-L1, kunye nokufunda nzulu.
Incwadi ibhalwe ngolwimi oluqhelekileyo, olufikelelekayo, oluqulethe ikhowudi ye-pseudo yee-algorithms eziphambili ezibalulekileyo.
Imixholo equkunjelwe kule ncwadi
- Unokwenzeka
- U kufunda o lukhulu
- Ukulungiswa rhoqo kweL1
- Ukulungiselelwa
- Ukulungiswa kombhalo
- izicelo zeKhompyutha Vision
- Ukusetyenziswa kweerobhothi
11. Imiba yokuFunda ngoBalo
Ngesakhelo sengqikelelo kunye nezifundo ezahlukeneyo, le ncwadi yokufunda ngomatshini ihlala yamkelwa ebaleni.
Le ncwadi inokusetyenziswa njengesalathiso kuye nabani na ofuna ukuxukuxa kwizihloko ezinje ngothungelwano lwe-neural kunye neendlela zokuvavanya kunye nentshayelelo elula yokufunda koomatshini.
Incwadi ityhala ngokungqongqo umfundi ukuba enze ezabo iimfuniselo kunye nophando ngalo lonke ixesha, iyenza ixabiseke ekukhuliseni izakhono kunye nomdla ofunekayo wokwenza inkqubela phambili kubuchule bokufunda kumatshini okanye umsebenzi.
Sisixhobo esibalulekileyo sabasebenzi beenkcukacha-manani kunye naye nabani na onomdla kwimigodi yedatha kushishino okanye kwisayensi. Qinisekisa ukuba uyayiqonda ialgebra yomgca ubuncinane phambi kokuba uqalise le ncwadi.
Imixholo equkunjelwe kule ncwadi
- Ukufunda okubekwe esweni (ukuqikelelwa) ukuya kwimfundo engajongwanga
- Unxibelelwano lweeNeural
- Oomatshini bokuxhasa intsholongwane
- Imithi yokuhlela
- Ukunyusa algorithms
12. UkuNakwa kwePatheni kunye nokuFunda koomatshini
Ihlabathi lokuqatshelwa kwepateni kunye nokufunda koomatshini kunokuphononongwa ngokucokisekileyo kule ncwadi. Indlela yeBayesi yokuqatshelwa kwepateni yaboniswa kolu papasho.
Ngapha koko, le ncwadi iphonononga izifundo ezinomngeni ezifuna ukuqonda okusebenzayo kwe-multivariate, isayensi yedatha, kunye ne-algebra esisiseko.
Ngokufunda koomatshini kunye nokuba nokwenzeka, incwadi yereferensi ibonelela ngezahluko ezinamanqanaba aqhubekayo anzima okuntsonkotha okusekwe kwiintsingiselo kwiiseti zedatha. Imizekelo elula inikwa phambi kwentshayelelo jikelele yokuqatshelwa kwepateni.
Le ncwadi ibonelela ngobuchule bokuqikelela okuqikelelweyo, okuvumela uqikelelo olukhawulezayo kwiimeko xa izisombululo ezichanekileyo zingenakwenzeka. Azikho ezinye iincwadi ezisebenzisa iimodeli zegraphical ukuchaza ukuhanjiswa okunokwenzeka, kodwa kunjalo.
Imixholo equkunjelwe kule ncwadi
- iindlela Bayesian
- Ii-algorithms eziqikelelweyo zokuthelekelela
- Iimodeli ezintsha ezisekwe kwiinkozo
- Intshayelelo yethiyori enokwenzeka esisiseko
- Intshayelelo yokuqatshelwa kwepateni kunye nokufunda koomatshini
13. Iziseko zokuFunda koomatshini kuHlalutyo lweDatha eQhelekileyo
Ukuba uye wazi kakuhle iziseko zokufunda koomatshini kwaye ufuna ukuqhubela phambili kuhlalutyo lwedatha eqikelelweyo, le yincwadi yakho !!! Ngokufumana iipatheni kwiiseti zedatha ezinkulu, ukuFunda ngoomatshini kunokusetyenziselwa ukuphuhlisa iimodeli zokuqikelela.
Le ncwadi iphonononga ukuphunyezwa kokusetyenziswa kweML Uhlalutyo lwedatha oluqikelelweyo nzulu, kuquka yomibini imigaqo yethiyori kunye nemizekelo eyiyo.
Ngaphandle kwento yokuba isihloko esithi "IiNqobo zokuFunda koMshini wokuHlalutywa kweDatha yokuQala" ngumlomo, le ncwadi iya kuchaza uhambo lwe-Predictive Data Analytics ukusuka kwidatha ukuya kwingqiqo ukuya kwisigqibo.
Ikwaxoxa ngeendlela ezine zokufunda koomatshini: ukufunda okusekwe kulwazi, ukufunda okusekwe ngokufana, ukufunda okunokwenzeka, kunye nokufunda okusekwe kwimpazamo, nganye inenkcazo engeyiyo yobugcisa elandelwa yimodeli yezibalo kunye ne-algorithms enemizekelo.
Imixholo Egutyungelwe encwadini
- Ukufunda okusekwe kulwazi
- Ukufunda okusekwe ngokufana
- Ukufunda okusekwe ekunokwenzeka
- Ukufunda okusekwe kwiimpazamo
14. IModeli yokuQikelela eSetyenzisiweyo
I-Applied Predictive Modeling iphonononga yonke inkqubo yokuqikelela kwangaphambili, iqala ngamanqanaba abalulekileyo okwenziwa kwedatha kwangaphambili, ukwahlulwa kwedatha, kunye nesiseko sokulungisa imodeli.
Umsebenzi ke ubonisa iinkcazo ezicacileyo zeendlela ezahlukeneyo eziqhelekileyo kunye nokutsha kunye neendlela zokuhlela, ngokugxininisa ekuboniseni nasekusombululeni imingeni yedatha yangempela.
Isikhokelo sibonisa yonke imiba yenkqubo yokwenza imodeli enezandla ezininzi, imizekelo yehlabathi langempela, kwaye isahluko ngasinye siquka ikhowudi ye-R ebanzi kwinqanaba ngalinye lenkqubo.
Lo mthamo wezinto ezininzi ungasetyenziswa njengentshayelelo kwiimodeli ezixelwe kwangaphambili kunye nenkqubo yonke yokumisela, njengesikhokelo sokubhekisa kubasebenzi, okanye njengesicatshulwa kwizifundo zokuqikelela inqanaba lesidanga sokuqala.
Imixholo equkunjelwe kule ncwadi
- Indlela yokubuyisela umva
- Ubuchule bokuhlela
- Ii-algorithms zeML ezintsonkothileyo
15. Ukufunda ngoomatshini: UbuGcisa kunye neNzululwazi yee-algorithms ezenza iSense yeDatha
Ukuba ungumntu ophakathi okanye oyingcali yokufunda koomatshini kwaye ufuna ukubuyela "kwizinto ezisisiseko," le ncwadi yeyakho! Ihlawula ngetyala elipheleleyo kumatshini wokuntsonkotha nobunzulu obumangalisayo ngelixa ungaze uphulukane nombono wemigaqo yokumanyanisa (impumelelo enkulu!).
Ukufunda ngoomatshini: UbuGcisa kunye neNzululwazi yee-algorithms zibandakanya izifundo ezininzi zobunzima obukhulayo, kunye nemizekelo emininzi kunye nemifanekiso (ukugcina izinto ezinomdla!).
Le ncwadi ikwaquka uluhlu olubanzi lweemodeli ezinengqondo, zejometri, kunye nezibalo, kunye nezifundo ezintsonkothileyo kunye nenoveli ezifana ne-matrix factorization kunye nohlalutyo lwe-ROC.
Imixholo equkunjelwe kule ncwadi
- Yenza lula i-algorithms yokufunda koomatshini
- Imodeli esengqiqweni
- Imodeli yejometri
- Imodeli yeenkcukacha-manani
- Uhlalutyo lwe-ROC
16. Ukumbiwa kweenkcukacha: IziXhobo zokuFunda zoomatshini kunye nobuGcisa
Ukusebenzisa iindlela ezisuka kufundo lweenkqubo zedathabheyisi, ukufundwa koomatshini, kunye nezibalo, ubuchule bokumbiwa kwedatha busenza sikwazi ukufumana iipatheni kwiimali ezininzi zedatha.
Ufanele ufumane incwadi yeDatha yeMigodi: IziXhobo zokuFunda ngoMatshini oSebenzayo kunye nobuChwephesha ukuba ufuna ukufunda ubuchule bokumbiwa kwedatha ngokukodwa okanye ucwangcise ukufunda umatshini wokufunda ngokubanzi.
Eyona ncwadi ibalaseleyo yokufunda koomatshini igxile ngakumbi kwicala layo lobugcisa. Ijonge ngakumbi kubuchwephesha bobugcisa bokufunda komatshini, kunye nezicwangciso zokuqokelela idatha kunye nokusebenzisa amagalelo neziphumo ezahlukeneyo ukugweba iziphumo.
Imixholo equkunjelwe kule ncwadi
- Iimodeli zomgca
- Clustering
- Ukwenziwa kweenkcukacha-manani
- Ukuqikelela ukusebenza
- Ukuthelekisa iindlela zokumbiwa kwedatha
- Ukufunda okusekwe kwimeko
- Ukumelwa kolwazi & namaqela
- Ubuchwephesha bemveli kunye nanamhlanje bokumbiwa kwedatha
17. IPython yoHlalutyo lweDatha
Ukukwazi ukuvavanya idatha esetyenziswa ekufundeni ngomatshini sesona buchule sibalulekileyo isazi sedatha ekufuneka sibe naso. Ngaphambi kokuphuhlisa imodeli ye-ML evelisa uqikelelo oluchanekileyo, uninzi lomsebenzi wakho luya kubandakanya ukuphatha, ukucubungula, ukucocwa, kunye nokuvavanya idatha.
Kufuneka uqhelane neelwimi zeprogram ezifana nePandas, iNumPy, i-Ipython, kunye nezinye ukuze wenze uhlalutyo lwedatha.
Ukuba ufuna ukusebenza kwisayensi yedatha okanye ukufundwa komatshini, kufuneka ube nobuchule bokusebenzisa idatha.
Ngokuqinisekileyo kuya kufuneka ufunde incwadi yePython yoHlahlelo lweDatha kule meko.
Imixholo equkunjelwe kule ncwadi
- Kubalulekile Amathala eencwadi ePython
- IiPanda eziPhambili
- Imizekelo yoHlahlelo lweDatha
- Ukucocwa kweDatha kunye nokuLungisa
- Iindlela zeMathematika nezoBalo
- Ushwankathelo kunye neComputing Statistics ezichazayo
18. UkuLungiswa koLwimi lweNdalo ngePython
Isiseko seenkqubo zokufunda koomatshini kukusetyenzwa kolwimi lwendalo.
Incwadi ethi Natural Language Processing with Python ikuyala ngendlela yokusebenzisa i-NLTK, ingqokelela ethandwa kakhulu yeemodyuli zePython kunye nezixhobo zokusetyenzwa kolwimi lwendalo olungokomfuziselo kunye nezibalo zesiNgesi kunye ne-NLP ngokubanzi.
I-Natural Language Processing kunye nencwadi yePython ibonelela ngeendlela ezisebenzayo zePython ezibonisa i-NLP ngendlela emfutshane, ecacileyo.
Abafundi banokufikelela kwiiseti zedatha ezichazwe kakuhle zokujongana nedatha engacwangciswanga, ulwakhiwo lolwimi lwesicatshulwa, kunye nezinye izinto ezigxile kwi-NLP.
Imixholo equkunjelwe kule ncwadi
- Lusebenza njani ulwimi lwabantu?
- Ulwakhiwo lwedatha yolwimi
- Izixhobo zoLwimi lweNdalo (NLTK)
- Ukucazulula kunye nohlalutyo lwesemantic
- Oovimba beenkcukacha zolwimi ezidumileyo
- Dibanisa ubuchule ukusuka kukubhadla okungeyonyani kunye neelwimi
19. Inkqubo yoBukrelekrele obuHlanganisiweyo
I-Programming Collective Intelligence nguToby Segaran, ethathwa njengenye yeencwadi ezinkulu zokuqalisa ukuqonda ukufundwa koomatshini, yabhalwa kwi-2007, iminyaka ngaphambi kokuba isayensi yedatha kunye nokufunda koomatshini kufumane indawo yabo yangoku njengeendlela ezihamba phambili zobuchwephesha.
Incwadi isebenzisa iPython njengendlela yokusasaza ubuchule bayo kubaphulaphuli bayo. Ubukrelekrele obuHlanganisiweyo beNkqubo bungaphezulu kwencwadana yokuphunyezwa kwe ml kunokuba iyintshayelelo yokufunda koomatshini.
Incwadi ibonelela ngolwazi ekuphuhliseni i-algorithms ye-ML esebenzayo yokuqokelela idatha kwii-apps, inkqubo yokufumana idatha kwiiwebhusayithi, kunye nokukhupha idatha eqokelelweyo.
Isahluko ngasinye siquka imisebenzi yokwandisa ii-algorithms ezixoxiwe kunye nokuphucula ukusebenziseka kwazo.
Imixholo equkunjelwe kule ncwadi
- Uhluzo lwaseBayesi
- Oomatshini bokuxhasa intsholongwane
- I-algorithms ye-injini yokukhangela
- Iindlela zokwenza uqikelelo
- Iindlela zokucoca ezisebenzisanayo
- I-non-negative matrix factorization
- Ubukrelekrele obuphuhlayo bokusombulula iingxaki
- Iindlela zokubona amaqela okanye iipateni
20. UkuFunda nzulu (i-Adaptive Computation kunye nochungechunge lokuFunda ngoomatshini)
Njengoko sonke sisazi, ukufunda okunzulu luhlobo oluphuculweyo lokufunda koomatshini olwenza ukuba iikhompyuter zifunde ekusebenzeni kwangaphambili kunye nomthamo omkhulu wedatha.
Ngelixa usebenzisa ubuchule bokufunda ngoomatshini, kuya kufuneka unxibelelane nemigaqo yokufunda enzulu. Le ncwadi, egqalwa njengeBhayibhile yemfundo enzulu, iya kuba luncedo kakhulu kule meko.
Iingcali ezintathu zokufunda nzulu zigubungela izihloko ezintsonkothileyo ezizaliswe yimathematika kunye nemizekelo enzulu yemveliso kule ncwadi.
Ukubonelela ngesiseko semathematika kunye nengqikelelo, umsebenzi uxoxa ngezimvo ezifanelekileyo kwialjebra yomgca, ithiyori enokwenzeka, ithiyori yolwazi, ukubalwa kwamanani, kunye nokufunda koomatshini.
Ivavanya usetyenziso olunjengokusetyenzwa kolwimi lwendalo, ukuqondwa kwentetho, umbono wekhompyuter, iinkqubo zengcebiso kwi-intanethi, i-bioinformatics, kunye nemidlalo yevidiyo kwaye ichaza ubuchule bokufunda obunzulu obusetyenziswa ngabasebenzi boshishino, njengothungelwano olunzulu lwe-feedforward, uhlengahlengiso, kunye ne-algorithms yokuphucula, uthungelwano lwe-convolutional, kunye neendlela ezisebenzayo. .
Imixholo equkunjelwe kule ncwadi
- Ubalo lwamanani
- UPhando lokuFunda olunzulu
- Iindlela zeComputer Vision
- Uthungelwano lweNgxelo enzulu
- UPhuculo lweMifuziselo eNzulu yoQeqesho
- Indlela eSebenzayo
- UPhando lokuFunda olunzulu
isiphelo
Iincwadi ezingama-20 eziphezulu zokufunda koomatshini zishwankathelwa kolu luhlu, onokulusebenzisa ukuqhubela phambili ukufunda ngomatshini kwicala olithandayo.
Uya kuba nakho ukuphuhlisa isiseko esiluqilima kubuchwephesha bokufunda koomatshini kunye nelayibrari yesalathiso onokuyisebenzisa rhoqo ngelixa usebenza kwindawo ukuba ufunda iindidi zezi ncwadi zezifundo.
Uya kukhuthazwa ukuba uqhubeke ufunda, usiba ngcono, kwaye ube nefuthe nokuba ufunda incwadi enye.
Xa ulungele kwaye unobuchule bokuphuhlisa eyakho ialgorithms yokufunda koomatshini, hlala ukhumbula ukuba idatha ibaluleke kakhulu kwimpumelelo yeprojekthi yakho.
Shiya iMpendulo