Data Science bakeng sa ba qalang
1. Tlhahlobo ea maikutlo (Tlhahlobo ea maikutlo ka mongolo)
Sheba phethahatso e felletseng ea projeke ea Data Science u sebelisa khoutu ea mohloli − .
Sentiment Analysis ke tlhahlobisiso ea mantsoe ho hlwaya maikutlo le maikutlo, a ka bang ntle kapa a fosahetse. Ona ke mofuta oa likarolo moo lihlopha li ka bang tsa binary (tse ntle le tse mpe) kapa tse ngata (tse thabileng, tse halefileng, tse hlomohileng, tse mpe...). Re tla kenya ts'ebetsong morero ona oa Data Science ho R 'me re tla sebelisa dataset ho sephutheloana sa "janeaustenR". Re tla sebelisa didikishinari tsa sepheo se akaretsang joalo ka AFINN, bing le loughran, re etsa kopanelo ea kahare mme qetellong re tla theha leru la mantsoe ho bonts'a sephetho.
Язык: R
Sephutheloana sa Boitsebiso/Sephutheloana: janeoustenR
Sengoloa se fetoletsoe ka tšehetso ea EDISON Software, e , hammoho le .
2. Fake News fumanoa
Isa tsebo ea hau boemong bo latelang ka ho sebetsa ho Data Science Project for Beginners − .
Litaba tsa bohata ke boitsebiso ba bohata bo phatlalatsoang ka mecha ea litaba ea sechaba le mecha e meng ea marang-rang e le ho finyella lipakane tsa lipolotiki. Mohopolong ona oa morero oa Data Science, re tla sebelisa Python ho aha mohlala o ka tsebang hantle hore na litaba ke tsa 'nete kapa ke tsa bohata. Re tla theha TfidfVectorizer mme re sebelise PassiveAggressiveClassifier ho hlophisa litaba ka "nete" le "fake". Re tla sebelisa dataset ea sebopeho sa 7796 × 4 mme re etse tsohle ho Jupyter Lab.
Язык: python
Sephutheloana sa Boitsebiso/Sephutheloana: litaba.csv
3. Ho lemoha lefu la Parkinson
Tsoela pele ka ho sebetsa ka Morero oa Tlhahiso ea Data Science Project − .
Re se re qalile ho sebelisa Data Science ho ntlafatsa tlhokomelo ea bophelo le litšebeletso - haeba re ka bolela esale pele lefu lena ka nako e sa le teng, re tla ba le melemo e mengata. Kahoo, mohopolong ona oa projeke ea Data Science, re tla ithuta ho lemoha lefu la Parkinson re sebelisa Python. Ke lefu la neurodegenerative, le tsoelang pele la tsamaiso ea methapo e bohareng le amang ho sisinyeha le ho baka ho thothomela le ho satalla. E ama li-neurone tse hlahisang dopamine bokong, 'me selemo se seng le se seng, e ama batho ba fetang limilione tse 1 India.
Язык: python
Sephutheloana sa Boitsebiso/Sephutheloana: Lethathamo la lintlha tsa UCI ML Parkinsons
Merero ea Saense ea data ea ho rarahana ha mahareng
4. Keletso ea Maikutlo a Puo
Sheba ho kengoa tšebetsong ka botlalo ha projeke ea sampole ea Data Science - .
Ha re ithuteng ho sebelisa lilaebrari tse fapaneng. Morero ona oa Mahlale a Boitsebiso o sebelisa librosa bakeng sa temoho ea puo. SER ke mokhoa oa ho khetholla maikutlo a batho le maemo a amehang ho tsoa puong. Hobane re sebelisa molumo le semelo ho hlahisa maikutlo ka mantsoe a rona, SER e bohlokoa. Empa kaha maikutlo a na le moelelo, tlhaloso ea molumo ke mosebetsi o boima. Re tla sebelisa mesebetsi ea mfcc, chroma le mel mme re sebelise dataset ea RAVDESS bakeng sa temoho ea maikutlo. Re tla theha sehlopha sa MLPC sa mofuta ona.
Язык: python
Sephutheloana sa Boitsebiso/Sephutheloana: Setšoantšo sa RAVDESS
5. Ho lemoha Tekano le Lilemo
Khahlisa bahiri ka morero oa morao-rao oa Data Science - .
Ena ke Science Science e khahlisang e nang le Python. U sebelisa setšoantšo se le seng feela, u tla ithuta ho bolela esale pele bong le lilemo tsa motho. Ho sena, re tla u tsebisa ka Computer Vision le melao-motheo ea eona. Re tla haha 'me e tla sebelisa mehlala e koetlisitsoeng ke Tal Hassner le Gil Levy ho Adience dataset. Re tla sebelisa lifaele tse ling tsa .pb, .pbtxt, .prototxt le .caffemodel tseleng.
Язык: python
Sephutheloana sa Boitsebiso/Sephutheloana: Adience
6. Tlhahlobo ea Boitsebiso ba Uber
Sheba phethahatso e felletseng ea projeke ea Data Science ka khoutu ea mohloli − .
Ona ke projeke ea pono ea data e nang le ggplot2 eo ho eona re tla sebelisa R le lilaebrari tsa eona le ho sekaseka mekhahlelo e fapaneng. Re tla sebelisa pokello ea lintlha tsa Uber Pickups New York le ho etsa lipono bakeng sa liforeimi tse fapaneng tsa nako tsa selemo. Sena se re bolella ka moo nako e amang maeto a bareki.
Язык: R
Sephutheloana sa Boitsebiso/Sephutheloana: Uber Pickups sebakeng sa datha sa New York City
7. Ho lemoha ho otsela ha Mokhanni
Ntlafatsa tsebo ea hau ka ho sebetsa ho Top Data Science Project - .
Ho khanna ka boroko ho kotsi haholo, 'me ho na le likotsi tse ka bang sekete selemo se seng le se seng ka lebaka la bakhanni ba robalang ba ntse ba khanna. Morerong ona oa Python, re tla theha sistimi e ka lemohang bakhanni ba robetseng le ho ba hlokomelisa ka lerata.
Morero ona o kengoa ts'ebetsong ka Keras le OpenCV. Re tla sebelisa OpenCV ho bona sefahleho le mahlo 'me ka thuso ea Keras re tla arola boemo ba mahlo (O butsoe kapa o Koetsoe) re sebelisa mekhoa e tebileng ea neural network.
8.Chatbot
Theha chatbot le Python 'me u hatele pele mosebetsing oa hau - .
Li-chatbots ke karolo ea bohlokoa ea khoebo. Likhoebo tse ngata li tlameha ho fana ka lits'ebeletso ho bareki ba tsona mme ho hloka matla a mangata, nako le matla ho li sebeletsa. Li-chatbots li ka iketsetsa boholo ba tšebelisano ea bareki ka ho araba tse ling tsa lipotso tse tloaelehileng tseo bareki ba li botsang. Ho na le mefuta e 'meli ea li-chatbots: Domain-specific le Open-domain. Chatbot e khethehileng ea domain hangata e sebelisoa ho rarolla bothata bo itseng. Kahoo, o hloka ho e etsa ka mokhoa o ikhethileng hore e sebetse hantle tšimong ea hau. Li-chatbots tse bulehileng li ka botsoa lipotso leha e le life, kahoo ho li koetlisa ho hloka boitsebiso bo bongata.
Sehlopha sa lintlha: Intents json file
Язык: python
Merero ea Advanced Data Science
9. Image Caption Generator
Sheba phethahatso ea morero ka khoutu ea mohloli - .
Ho hlalosa se setšoantšong ke mosebetsi o bonolo ho batho, empa bakeng sa likhomphutha, setšoantšo ke pokello ea linomoro tse emelang boleng ba 'mala oa pixel ka 'ngoe. Ona ke mosebetsi o boima bakeng sa likhomphutha. Ho utloisisa se leng setšoantšong le ho theha tlhaloso ea puo ea tlhaho (mohlala, Senyesemane) ke mosebetsi o mong o boima. Morero ona o sebelisa mekhoa e tebileng ea ho ithuta moo re kenyang tšebetsong Convolutional Neural Network (CNN) ka Recurrent Neural Network (LSTM) ho theha jenereithara ea tlhaloso ea setšoantšo.
Sehlopha sa lintlha: Flickr 8K
Язык: python
Moralo: Keras
10. Ho Fumana Bomenemene ka Karete ea Mokitlane
Etsa sohle se matleng a hau ka ho sebetsa mohopolong oa projeke ea Data Science − .
Hona joale u se u qalile ho utloisisa mekhoa le likhopolo. Ha re feteleng ho merero e tsoetseng pele ea mahlale a data. Morerong ona, re tla sebelisa puo ea R e nang le li-algorithms tse kang , ho fokotseha ha lintho, marang-rang a maiketsetso a methapo ea kutlo le sehlopha sa ho eketsa sekhahla. Re tla sebelisa pokello ea lintlha tsa karete ea mokoloto ho khetholla litšebelisano tsa likarete tsa mokoloto e le tsa bomenemene le tsa 'nete. Re tla khetha mefuta e fapaneng bakeng sa bona mme re hahe li-curve tsa ts'ebetso.
Язык: R
Sephutheloana sa Boitsebiso/Sephutheloana: Lethathamo la lintlha tsa likarete tsa Transactions
11. Movie Keletso System
Lekola ts'ebetsong ea projeke e ntle ka ho fetisisa ea Data Science ka Source Code -
Morerong ona oa Mahlale a Boitsebiso, re tla sebelisa R ho phethahatsa likhothaletso tsa filimi ka ho ithuta ka mochini. Sistimi ea likhothaletso e romela litlhahiso ho basebelisi ka mokhoa oa ho sefa o ipapisitse le likhetho tsa basebelisi ba bang le nalane ea ho bala. Haeba A le B ba rata Home Alone, 'me B ba rata Mean Girls, u ka etsa tlhahiso ea A - le bona ba ka e rata. Sena se lumella bareki ho sebelisana le sethala.
Язык: R
Sephutheloana sa Boitsebiso/Sephutheloana: Sehlopha sa data sa MovieLens
12. Karohano ea Bareki
Khahlisa bahiri ka projeke ea Data Science (ho kenyeletsoa le khoutu ea mohloli) - .
Karolo ea bareki ke sesebelisoa se tsebahalang . Ka ho sebelisa lihlopha, lik'hamphani li hlalosa likarolo tsa bareki ho sebetsa le basebetsi ba ka bang teng. Ba arola bareki ka lihlopha ho latela litšobotsi tse tloaelehileng tse kang bong, lilemo, lithahasello le mekhoa ea ho sebelisa chelete, e le hore ba ka bapatsa ka katleho lihlahisoa tsa bona sehlopheng ka seng. Re tla sebelisa , hammoho le ho bona ka mahlo a kelello kabo ea thobalano le lilemo. Ebe re sekaseka letseno la bona la selemo le maemo a tšebeliso ea chelete.
Язык: R
Sephutheloana sa Boitsebiso/Sephutheloana: Mall_Customers pokello ea lintlha
13. Sehlopha sa Kankere ea Matsoele
Bona ho kengoa ts'ebetsong ka botlalo ha morero oa Data Science ho Python - .
Ha re khutlela ho tlatsetso ea bongaka ea mahlale a data, ha re ithuteng ho lemoha mofetše oa matsoele ka Python. Re tla sebelisa IDC_regular dataset ho bona invasive ductal carcinoma, mofuta o atileng haholo oa mofetše oa matsoele. E hlaha ka har'a likotopo tsa lebese, e kenelle ka har'a mahlahahlaha kapa mafura a tšoelesa ea mammary ka ntle ho lesela. Ka mohopolo ona oa morero oa pokello ea data, re tla o sebelisa le laebrari ea Keras bakeng sa ho aroloa.
Язык: python
Sephutheloana sa Boitsebiso/Sephutheloana: IDC_e tloaelehileng
14. Tlhokomeliso ea Matšoao a Sephethephethe
Ho fihlella ho nepahala ha thekenoloji ea likoloi tsa ho itsamaisa ka morero oa Data Science ka mohloli o bulehileng.
Matšoao a litsela le melao ea sephethephethe li bohlokoa haholo ho mokhanni e mong le e mong ho qoba likotsi. Ho latela molao, u lokela ho qala ka ho utloisisa hore na letšoao la tsela le shebahala joang. Motho o tlameha ho ithuta matšoao ohle a tsela pele a fuoa tokelo ea ho khanna koloi leha e le efe. Empa hona joale palo ea likoloi tse ikemetseng e ntse e eketseha, 'me haufinyane motho a ke ke a hlola a khanna koloi a le mong. Lenaneong la Tsebiso ea Letšoao la Tsela, u tla ithuta kamoo lenaneo le ka lemohang mofuta oa letšoao la tsela ka ho nka setšoantšo e le ho kenya letsoho. Jeremane Road Sign Recognition Reference Dataset (GTSRB) e sebelisetsoa ho haha marang-rang a tebileng a methapo ea kutlo ho lemoha sehlopha seo letšoao la sephethephethe le leng ho sona. Re boetse re theha GUI e bonolo bakeng sa ho sebelisana le ts'ebeliso.
Язык: python
Sehlopha sa lintlha: GTRB (Letshwao la Kamohelo ya Letshwao la Sephethephethe sa Jeremane)
Bala haholoanyane
Source: www.habr.com