Mapulojekiti 14 otseguka kuti mukweze luso lanu la Science Science (zosavuta, zachilendo, zolimba)

Sayansi ya Data kwa Oyamba

1. Kusanthula Maganizo (Kusanthula Maganizo Kupyolera M'malemba)

Onani kukhazikitsidwa kwathunthu kwa projekiti ya Data Science pogwiritsa ntchito ma source code − Sentiment Analysis Project mu R.

Sentiment Analysis ndi kusanthula kwa mawu kuti mudziwe malingaliro ndi malingaliro, omwe angakhale abwino kapena oipa. Uwu ndi mtundu wamagulu momwe makalasi amatha kukhala apawiri (zabwino ndi zoyipa) kapena kuchuluka (osangalala, okwiya, achisoni, oyipa ...). Tidzagwiritsa ntchito pulojekitiyi ya Data Science mu R ndipo tidzagwiritsa ntchito deta mu phukusi la "janeaustenR". Tidzagwiritsa ntchito madikishonale monga AFINN, bing ndi loughran, lowetsani mkati, ndipo pamapeto tidzapanga mtambo wa mawu kuti muwonetse zotsatira.

Chilankhulo: R
Seti / Phukusi: JaneaustenR

Nkhaniyi idamasuliridwa mothandizidwa ndi EDISON Software, yomwe amapanga zipinda zofananirako zamashopu amitundu yambirindipo mapulogalamu mapulogalamu.

2. Kuzindikira Nkhani Zabodza

Tengani luso lanu pamlingo wina pogwira ntchito ya Data Science kwa oyamba kumene - kupeza nkhani zabodza ndi Python.

Nkhani zabodza ndi zabodza zomwe zimafalitsidwa kudzera pawailesi yakanema komanso ma TV ena kuti akwaniritse zolinga zandale. Mu lingaliro la polojekiti ya Data Science, tidzagwiritsa ntchito Python kupanga chitsanzo chomwe chingadziwe bwino ngati nkhani yankhani ndi yeniyeni kapena yabodza. Tidzapanga TfidfVectorizer ndikugwiritsa ntchito PassiveAggressiveClassifier kugawa nkhani kukhala "zenizeni" ndi "zabodza". Tidzagwiritsa ntchito deta ya mawonekedwe 7796 × 4 ndikuyendetsa chirichonse mu Jupyter Lab.

Chilankhulo: Python

Seti / Phukusi: nkhani.csv

3. Kuzindikira Matenda a Parkinson

Pitirizani patsogolo ndi Idea yanu ya Data Science Project - kuzindikira matenda a Parkinson pogwiritsa ntchito XGBoost.

Tayamba kugwiritsa ntchito Data Science kupititsa patsogolo chisamaliro chaumoyo ndi mautumiki - ngati tingathe kulosera za matenda adakali aang'ono, ndiye kuti tidzakhala ndi zabwino zambiri. Chifukwa chake, mu lingaliro la polojekiti ya Data Science, tiphunzira momwe tingadziwire matenda a Parkinson pogwiritsa ntchito Python. Ndi matenda a neurodegenerative, opita patsogolo apakati pa mitsempha yapakati yomwe imakhudza kuyenda ndikuyambitsa kunjenjemera ndi kuuma. Zimakhudza ma neuroni omwe amapanga dopamine mu ubongo, ndipo chaka chilichonse, zimakhudza anthu oposa 1 miliyoni ku India.

Chilankhulo: Python

Seti / Phukusi: Zithunzi za UCI ML Parkinsons

Ma projekiti a Science Science apakati zovuta

4. Kuzindikira Kutengeka kwa Mawu

Onani kukhazikitsidwa kwathunthu kwa projekiti yachitsanzo ya Data Science − kuzindikira mawu pogwiritsa ntchito Librosa.

Tsopano tiyeni tiphunzire mmene tingagwiritsire ntchito malaibulale osiyanasiyana. Pulojekiti ya Data Science iyi imagwiritsa ntchito librosa pozindikira mawu. SER ndi njira yozindikiritsira momwe anthu akumvera komanso momwe amamvera polankhula. Popeza timagwiritsa ntchito mawu ndi mawu kuti tifotokoze zakukhosi ndi mawu athu, SER ndiyofunikira. Koma popeza kutengeka mtima kumakhala kokhazikika, kutanthauzira mawu ndi ntchito yovuta. Tidzagwiritsa ntchito mfcc, chroma ndi mel ndikugwiritsa ntchito dataset ya RAVDESS kuti tizindikire zakukhosi. Tipanga gulu la MLPC lachitsanzochi.

Chilankhulo: Python

Seti / Phukusi: Zithunzi za RAVDESS

5. Kuzindikira Jenda ndi Zaka

Gwirani ntchito ndi projekiti yaposachedwa ya Data Science - kudziwa jenda ndi zaka pogwiritsa ntchito OpenCV.

Iyi ndi Science Science yosangalatsa yokhala ndi Python. Pogwiritsa ntchito chithunzi chimodzi, muphunzira kulosera za jenda ndi zaka za munthu. Mu izi tidzakudziwitsani za Computer Vision ndi mfundo zake. Tidzamanga convolutional neural network ndipo adzagwiritsa ntchito zitsanzo zophunzitsidwa ndi Tal Hassner ndi Gil Levy pa dataset ya Adience. Panjira tidzagwiritsa ntchito mafayilo ena a .pb, .pbtxt, .prototxt ndi .caffemodel.

Chilankhulo: Python

Seti / Phukusi: Adience

6. Uber Data Analysis

Onani kukhazikitsidwa kwathunthu kwa projekiti ya Data Science ndi code code − Uber Data Analysis Project mu R.

Iyi ndi ntchito yowonetsera deta ndi ggplot2 momwe tidzagwiritsira ntchito R ndi malaibulale ake ndikusanthula magawo osiyanasiyana. Tidzagwiritsa ntchito dataseti ya Uber Pickups ku New York City ndikupanga zowonera zanthawi zosiyanasiyana pachaka. Izi zikutiuza momwe nthawi imakhudzira maulendo a kasitomala.

Chilankhulo: R

Seti / Phukusi: Uber Pickups ku New York City dataset

7. Kuzindikira Kugona kwa Dalaivala

Limbikitsani luso lanu pogwira ntchito pa Top Data Science Project - Njira yodziwira kugona ndi OpenCV & Keras.

Kuyendetsa tulo ndi koopsa kwambiri, ndipo pafupifupi ngozi chikwi chimodzi zimachitika chaka chilichonse chifukwa cha madalaivala kugona akuyendetsa. Mu pulojekiti ya Python iyi, tipanga dongosolo lomwe limatha kuzindikira madalaivala akuwodzera komanso kuwachenjeza ndi mawu omvera.

Ntchitoyi ikugwiritsidwa ntchito pogwiritsa ntchito Keras ndi OpenCV. Tidzagwiritsa ntchito OpenCV pozindikira nkhope ndi maso ndipo ndi Keras tidzayika mawonekedwe amaso (Otsegula kapena Otsekedwa) pogwiritsa ntchito njira zakuya za neural network.

8. Wokambirana

Pangani Chatbot ndi Python ndikupita patsogolo pantchito yanu - Chatbot ndi NLTK & Keras.

Ma Chatbots ndi gawo lofunikira pabizinesi. Mabizinesi ambiri amayenera kupereka chithandizo kwa makasitomala awo ndipo zimatengera anthu ogwira ntchito, nthawi komanso khama kuti awatumikire. Ma Chatbots amatha kusintha zambiri zamakasitomala anu poyankha mafunso omwe makasitomala amafunsa. Pali mitundu iwiri ya ma chatbots: Domain-enieni ndi Open-domain. Chatbot yodziwika ndi domain nthawi zambiri imagwiritsidwa ntchito kuthetsa vuto linalake. Chifukwa chake, muyenera kuyisintha mwamakonda kuti igwire bwino ntchito m'munda wanu. Ma chatbots otseguka amatha kufunsidwa mafunso aliwonse, chifukwa chake kuwaphunzitsa kumafuna zambiri.

Seti ya data: Cholinga cha fayilo ya json

Chilankhulo: Python

Ntchito za Advanced Data Science

9. Image Caption Generator

Onani kukhazikitsidwa kwathunthu kwa projekitiyo ndi magwero a code - Image Caption Generator yokhala ndi CNN & LSTM.

Kufotokozera zomwe zili pachithunzichi ndi ntchito yosavuta kwa anthu, koma pamakompyuta, chithunzi chimangokhala manambala angapo omwe amayimira mtengo wamtundu wa pixel iliyonse. Iyi ndi ntchito yovuta kwa makompyuta. Kumvetsetsa zomwe zili m'chifaniziro ndikupanga kufotokozera m'chinenero chachibadwa (monga Chingerezi) ndi ntchito ina yovuta. Pulojekitiyi imagwiritsa ntchito njira zophunzirira mozama momwe timagwiritsira ntchito Convolutional Neural Network (CNN) yokhala ndi Recurrent Neural Network (LSTM) kuti tipange jenereta yofotokozera zithunzi.

Seti ya data: Chithunzi cha Flickr 8K

Chilankhulo: Python

Zomangamanga: Keras

10. Kuzindikira Chinyengo Cha Ngongole

Chitani zomwe mungathe pamene mukugwira ntchito pamalingaliro anu a projekiti ya Data Science - pezani chinyengo cha kirediti kadi pogwiritsa ntchito makina ophunzirira.

Pakalipano mwayamba kumvetsetsa njira ndi malingaliro. Tiyeni tipitirire kuzinthu zina zapamwamba za sayansi ya data. Mu polojekitiyi tidzagwiritsa ntchito chilankhulo cha R chokhala ndi ma aligorivimu ngati mitengo ya chisankho, kusinthika kwazinthu, maukonde opangira ma neural network ndi gradient boosting classifier. Tidzagwiritsa ntchito gulu lazochita zamakadi kuti tisankhe zochita pa kirediti kadi ngati zachinyengo kapena zenizeni. Tidzawasankhira mitundu yosiyanasiyana ndikupanga ma curve ochitira.

Chilankhulo: R

Seti / Phukusi: Zotengera za Khadi

11. Movie Malangizo System

Phunzirani kukhazikitsidwa kwa projekiti yabwino kwambiri ya Data Science ndi Source code - Kanema Malangizo System mu R chinenero

Mu pulojekiti ya Data Science iyi, tidzagwiritsa ntchito R kukhazikitsa malingaliro a kanemayu kudzera pophunzira pamakina. Dongosolo lothandizira limatumiza malingaliro kwa ogwiritsa ntchito kudzera muzosefera kutengera zomwe amakonda ogwiritsa ntchito ena komanso mbiri yosakatula. Ngati A ndi B amakonda Home Alone, ndipo B amakonda Mean Girls, mutha kupereka lingaliro A - angakondenso. Izi zimathandiza makasitomala kuyanjana ndi nsanja.

Chilankhulo: R

Seti / Phukusi: MovieLens dataset

12. Gawo la Makasitomala

Gwirani ntchito ndi projekiti ya Data Science (kuphatikiza ma code source) - Kugawa kwamakasitomala pogwiritsa ntchito kuphunzira pamakina.

Segmentation ya ogula ndi ntchito yotchuka maphunziro osayang'aniridwa. Pogwiritsa ntchito clustering, makampani amazindikira magawo amakasitomala kuti agwirizane ndi omwe angagwiritse ntchito. Amagawa makasitomala m'magulu molingana ndi mikhalidwe yodziwika bwino monga jenda, zaka, zokonda komanso momwe amawonongera ndalama kuti athe kugulitsa malonda awo ku gulu lililonse. Tidzagwiritsa ntchito K-amatanthauza masango, komanso kuwona m'maganizo kugawika kwa jenda ndi zaka. Kenako tidzasanthula ndalama zomwe amapeza pachaka komanso ndalama zomwe amawononga.

Chilankhulo: R

Seti / Phukusi: Mall_Customers dataset

13. Gulu la Khansa ya M'mawere

Onani kukhazikitsidwa kwathunthu kwa projekiti ya Data Science ku Python - Gulu la khansa ya m'mawere pogwiritsa ntchito kuphunzira mozama.

Kubwereranso ku chithandizo chachipatala cha sayansi ya data, tiyeni tiphunzire momwe tingadziwire khansa ya m'mawere pogwiritsa ntchito Python. Tigwiritsa ntchito IDC_regular dataset kuti tidziwe invasive ductal carcinoma, mtundu wofala kwambiri wa khansa ya m'mawere. Imakula m'mitsempha ya mkaka, ndikulowera mu minofu ya m'mawere ya ulusi kapena mafuta kunja kwa njirayo. Mu lingaliro la polojekiti yosonkhanitsira deta iyi tidzagwiritsa ntchito Kuphunzira Kwambiri ndi laibulale ya Keras kuti igawidwe.

Chilankhulo: Python

Seti / Phukusi: IDC_zokhazikika

14. Kuzindikira Zizindikiro Zamsewu

Kukwaniritsa zolondola paukadaulo wodziyendetsa nokha ndi projekiti ya Data Science kuzindikira chizindikiro cha traffic pogwiritsa ntchito CNN gwero lotseguka.

Zizindikiro zapamsewu ndi malamulo apamsewu ndizofunikira kwambiri kuti dalaivala aliyense apewe ngozi. Kuti muzitsatira lamuloli, choyamba muyenera kumvetsetsa momwe chizindikiro cha msewu chikuwonekera. Munthu ayenera kuphunzira zikwangwani zonse zamsewu asanapatsidwe laisensi yoyendetsa galimoto iliyonse. Koma tsopano chiwerengero cha magalimoto odziyimira pawokha chikukula, ndipo posachedwapa munthu sadzakhalanso kuyendetsa galimoto paokha. Mu pulojekiti Yozindikiritsa Chikwangwani Chamsewu, muphunzira momwe pulogalamu ingazindikire mtundu wa zikwangwani zamsewu potenga chithunzi ngati cholowetsa. Chidziwitso cha German Traffic Sign Recognition Benchmark (GTSRB) chimagwiritsidwa ntchito kupanga netiweki yakuya kuti izindikire kalasi yomwe chizindikirocho chili. Timapanganso GUI yosavuta kuti igwirizane ndi pulogalamuyi.

Chilankhulo: Python

Seti ya data: GTSRB (Chizindikiro Chozindikira Chizindikiritso cha Magalimoto ku Germany)

Werengani zambiri

• Ma dataset 52 a ntchito zophunzitsira
• Front-end dojo: mapulojekiti ophunzitsa luso la otukula (5 atsopano + 43 akale)
• Top 12 yosangalatsa kwambiri ya IT dynamic infographics

Source: www.habr.com