MECER MS-DP100T01 Ho Rala le ho Sebelisa Tharollo ea Saense ea Boitsebiso ka Azure
| DURATION | LEBAKA | THEKNOLOJI | PELO MOKHOA |
KHOETSO LIKELETE |
| 3 Matsatsi | Mahareng | Azure | E etelletsoe pele ke morupeli | NA |
LIEKETSENG
Fumana tsebo e hlokahalang mabapi le mokhoa oa ho sebelisa lits'ebeletso tsa Azure ho nts'etsapele, ho koetlisa le ho tsamaisa, tharollo ea ho ithuta ka mochini. Thupelo e qala ka ho qetelaview ea lits'ebeletso tsa Azure tse tšehetsang mahlale a data. Ho tloha moo, e tsepamisitse maikutlo ho sebeliseng ts'ebeletso ea pele ea mahlale a data ea Azure, ts'ebeletso ea ho Ithuta ea Mochini oa Azure, ho iketsetsa lipeipi tsa mahlale a data. Thupelo ena e tsepamisitse maikutlo ho Azure mme ha e rute moithuti mokhoa oa ho etsa mahlale a data. Ho nahanoa hore baithuti ba se ba ntse ba tseba seo.
BAMELELI PROFILE
Thupelo ena e etselitsoe bo-ramahlale ba data le ba nang le boikarabello ba bohlokoa ho koetlisa le ho tsamaisa mekhoa ea ho ithuta ka mochini.
LITS'ELISITSOE
Pele ba ea thupelong ena, baithuti ba tlameha ho ba le:
- Lintho tsa motheo tsa Azure
- Kutloisiso ea mahlale a data ho kenyelletsa mokhoa oa ho lokisa data, mehlala ea terene, le ho lekola mefuta e hlolisanoang ho khetha e ntle ka ho fetisisa.
- Mokhoa oa ho hlophisa ka puo ea lenaneo la Python le ho sebelisa lilaebrari tsa Python: pandas, scikit-learn, matplotlib, le seaborn.
TSEBISO TS'OANE
Kamora ho qeta thupelo ena, baithuti ba tla tseba ho:
- Utloisisa mahlale a data ho Azure
- Sebelisa ho Ithuta ka Mechini ho iketsetsa ts'ebetso ea ho qetela
- Laola le ho hlokomela tšebeletso ea ho Ithuta ka Mechini
Module 1: Ho Qala ka ho Ithuta ka Mochini oa Azure
Mojulung ona, o tla ithuta ho fana ka sebaka sa ho sebetsa sa Mochini oa Azure mme o se sebelise ho laola thepa ea ho ithuta ea mochini joalo ka data, compute, khoutu ea koetliso ea mohlala, metrics e kentsoeng, le mefuta e koetlisitsoeng. U tla ithuta mokhoa oa ho sebelisa web-e thehiloeng ho Azure Machine Learning studio interface hammoho le Azure Machine Learning SDK le lisebelisoa tsa nts'etsopele tse kang Visual Studio Code le Jupyter Notebooks ho sebetsa le thepa e sebakeng sa hau sa mosebetsi.
Lithuto
- Kenyelletso ea ho Ithuta ka Mochini oa Azure
- Ho sebetsa le ho Ithuta ka Mochini oa Azure
- Lab: Theha Sebaka sa ho Ithuta sa Mochini oa Azure
- Fana ka sebaka sa ho sebetsa sa Mochini oa Azure
- Sebelisa lisebelisoa le khoutu ho sebetsa le Azure Machine Learning
Module 2: Lisebelisoa tse Bonahalang tsa ho Ithuta ka Mochini
Mojule ona o hlahisa lisebelisoa tse bonoang tsa ho Ithuta ka Mochini le Moqapi, tseo o ka li sebelisang ho koetlisa, ho lekola, le ho tsamaisa mefuta ea ho ithuta ka mochini ntle le ho ngola khoutu efe kapa efe.
Lithuto
- Ithute ka Mechini
- Moqapi oa ho Ithuta oa Mochini oa Azure
Lab: Sebelisa Thuto ea Mochini e Ikemetseng
Lab: Sebelisa Moqapi oa ho Ithuta oa Mochini oa Azure
Kamora ho qeta mojule ona, o tla khona ho
- Sebelisa ho ithuta ka mochini ho koetlisa mohlala oa ho ithuta ka mochini
- Sebelisa moqapi oa ho ithuta oa Azure Machine ho koetlisa mohlala
Module 3: Ho Etsa Liteko le Mehlala ea Koetliso
Mojulung ona, o tla qala ka liteko tse kenyelletsang ts'ebetso ea data, khoutu ea koetliso ea mohlala, le ho li sebelisa ho koetlisa mefuta ea ho ithuta ea mochini. Lithuto
- Selelekela sa Liteko
- Mehlala ea Koetliso le Ngoliso
Lab: Mehlala ea Terene
Lab: Matha liteko
Kamora ho qeta mojule ona, o tla khona ho
- Etsa liteko tse thehiloeng ho khoutu sebakeng sa ho sebetsa sa Azure Machine
- Koetlisa le ho ngolisa mefuta ea ho ithuta ka mochini
Module 4: Ho sebetsa le Data Data
ke ntho ea bohlokoa mosebetsing ofe kapa ofe oa ho ithuta oa mochini, kahoo mojulung ona, o tla ithuta ho theha le ho laola polokelo ea data le li-database sebakeng sa mosebetsi sa Azure Machine Learning, le mokhoa oa ho li sebelisa litekong tsa koetliso ea mohlala.
Lithuto
- Ho sebetsa le Datastores
- Ho sebetsa le Datasets
Lab: Sebetsa le Data
Kamora ho qeta mojule ona, o tla khona ho
- Theha le ho sebelisa polokelo ea boitsebiso
- Theha le ho sebelisa li-database
Mojule 5: Ho sebetsa le Compute
E 'ngoe ea melemo ea mantlha ea leru ke bokhoni ba ho sebelisa lisebelisoa tsa komporo ha li batloa le ho li sebelisa ho lekanya lits'ebetso tsa ho ithuta ka mochini ho isa bohōleng bo ke keng ba khoneha ho lisebelisoa tsa hau. Mojulung ona, o tla ithuta ho laola maemo a liteko a netefatsang hore liteko li sebetsa ka mokhoa o tsitsitseng, le mokhoa oa ho theha le ho sebelisa lipehelo tsa khomphutha bakeng sa liteko.
Lithuto
- Ho sebetsa le Tikoloho
- Ho sebetsa le Compute Targets
Lab: Sebetsa le Compute
Kamora ho qeta mojule ona, o tla khona ho
- Theha le ho sebelisa tikoloho
- Theha le ho sebelisa lipheo tsa khomphutha
Mojule oa 6: Ho hlophisa Ts'ebetso ka Liphaephe
Kaha joale u utloisisa lintlha tsa motheo tsa ho tsamaisa mesebetsi e le liteko tse sebelisang matlotlo a data le lisebelisoa tsa khomphutha, ke nako ea ho ithuta ho hlophisa mesebetsi ena e le lipeipi tsa mehato e hokahaneng. Liphaephe ke senotlolo sa ho kenya tšebetsong tharollo e sebetsang ea Machine Learning Operationalization (ML Ops) ho Azure, kahoo o tla hlahloba mokhoa oa ho li hlalosa le ho li tsamaisa mojuleng ona.
Lithuto
- Selelekela ho Pipelines
- Phatlalatso le Liphaephe tse Tsamaisang
Lab: Etsa Pipeline
Kamora ho qeta mojule ona, o tla khona ho
- Theha liphaephe ho iketsetsa tšebetso ea ho ithuta ka mochini
- Phatlalatsa le ho tsamaisa litšebeletso tsa lipeipi
Mojule oa 7: Mehlala ea ho Beha le ho Sebelisa
Mefuta e etselitsoe ho thusa ho etsa liqeto ka likhakanyo, kahoo e thusa feela ha e rometsoe 'me e fumaneha hore kopo e sebelisoe. Mojulung ona ithute mokhoa oa ho kenya mehlala ea ho nyenyefatsa ka nako ea nnete, le bakeng sa ho etsa li-batch inferencing.
Lithuto
- Tšusumetso ea nako ea 'nete
- Batch Inferencing
- Kopanyo e Tsoelang Pele le Phano
Lab: Theha Ts'ebeletso ea Ts'ireletso ea Nako ea 'Nete
Lab: Theha Tšebeletso ea Batch Inferencing
Kamora ho qeta mojule ona, o tla khona ho
- Phatlalatsa mohlala e le ts'ebeletso ea boithuto ba nako ea nnete
- Phatlalatsa mohlala e le tšebeletso ea litšupiso tsa sehlopha
- Hlalosa mekhoa ea ho kenya ts'ebetsong ho kopanya ho tsoelang pele le ho fana
Mojule oa 8: Ho Koetlisa Mehlala e Ntle
Ka sena stagHa e le hantle, u ithutile mokhoa oa ho qetela oa koetliso, ho tsamaisa le ho sebelisa mekhoa ea ho ithuta ka mochine; empa u etsa bonnete ba hore mofuta oa hau o hlahisa liphetho tse ntle ka ho fetesisa tsa data ea hau? Mojuleng ona, o tla hlahloba hore na o ka sebelisa tokiso ea hyperparameter joang le ho ithuta ka mochini o ikemetseng ho nka pele.tage ea compute ea cloud-scale 'me u fumane mohlala o motle ka ho fetisisa oa data ea hau.
Lithuto
- Tokiso ea Hyperparameter
- Ithute ka Mechini
Lab: Sebelisa Thuto ea Mochini e Ikemetseng ho tsoa ho SDK
Lab: Tune Hyperparameters Kamora ho qeta mojule ona, o tla khona ho
- Ntlafatsa li-hyperparameter bakeng sa koetliso ea mohlala
- Sebelisa ho ithuta ka mochini ho fumana mofuta o nepahetseng bakeng sa data ea hau
Mojule oa 9: Boithuto ba Mochini o nang le Boikarabelo
Bo-rasaense ba datha ba na le mosebetsi oa ho etsa bonnete ba hore ba sekaseka lintlha le ho koetlisa mehlala ea ho ithuta mochini ka mokhoa o nang le boikarabelo; ho hlompha boinotši ba motho ka mong, ho fokotsa leeme, le ho netefatsa ponaletso. Mojule ona o hlahloba lintlha le mekhoa ea ho sebelisa melaoana e nang le boikarabelo ea ho ithuta ka mochini. Lithuto
- Boinotši bo fapaneng
- Tlhaloso ea Mohlala
- Ho loka
Lab: Lekola boitsebelo bo fapaneng
Lab: Hlalosa Mehlala
Lab: Lemoha le ho Fokotsa ho Hloka Toka Ka mor'a ho qeta mojule ona, u tla khona ho
- Sebelisa provacy e fapaneng ho tlhahlobo ea data
- Sebelisa litlhaloso ho toloka mefuta ea ho ithuta ka mochini
- Lekola mehlala bakeng sa toka
Module 10: Mehlala ea Tlhokomelo
Ka mor'a hore mohlala o sebelisoe, ho bohlokoa ho utloisisa hore na mohlala o sebelisoa joang tlhahisong, le ho lemoha tšenyo leha e le efe ea katleho ea eona ka lebaka la ho hoholeha ha data. Mojule ona o hlalosa mekhoa ea ho lekola mehlala le data ea eona. Lithuto
- Mehlala ea Tlhokomelo e nang le Maikutlo a Tšebeliso
- Tlhokomelo ea Data Drift
Lab: Hlahloba Data Drift
Lab: Lekola Mohlala o nang le Maikutlo a Tšebeliso
Kamora ho qeta mojule ona, o tla khona ho
- Sebelisa Maikutlo a Tšebeliso ho beha leihlo mohlala o phatlalalitsoeng
- Lekola phallo ea data
LIKETITIFICATION TSE TSOANG LE TLHAHLOBO
Thupelo ena e tla lokisetsa baemeli ho ngola Microsoft DP-100: Ho Rala le ho Kenya Tšebeliso ea Tharollo ea Saense ea Boitsebiso tlhatlhobong ea Azure.
Litokomane / Lisebelisoa
 |
MECER MS-DP100T01 Ho Rala le ho Sebelisa Tharollo ea Saense ea Boitsebiso ka Azure [pdf] Bukana ea Mosebelisi MS-DP100T01 Ho Rala le ho Kenyelletsa Tharollo ea Saense ea Boitsebiso ka Azure, MS-DP100T01, Ho Rala le ho Kenyelletsa Tharollo ea Saense ea Boitsebiso ka Azure |
