{"id":54096,"date":"2019-12-18T00:00:00","date_gmt":"2019-12-17T21:00:00","guid":{"rendered":"https:\/\/prohoster.info\/blog\/blog_prohoster\/52-dataseta-dlya-trenirovochnyh-proektov"},"modified":"2020-02-18T14:02:04","modified_gmt":"2020-02-18T11:02:04","slug":"52-dataseta-dlya-trenirovochnyh-proektov","status":"publish","type":"post","link":"https:\/\/prohoster.info\/hu\/blog\/news\/52-dataseta-dlya-trenirovochnyh-proektov","title":{"rendered":"52 adatgy\u0171jtem\u00e9ny edz\u0151projektekhez","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
    \n
  1. Bolt V\u00e1s\u00e1rl\u00f3i Adatb\u00e1zis<\/a><\/noindex> \u2014 bolt l\u00e1togat\u00f3k adatai: azonos\u00edt\u00f3, nem, \u00e9letkor, j\u00f6vedelem, k\u00f6lt\u00e9si index. (Alkalmaz\u00e1si lehet\u0151s\u00e9g: V\u00e1s\u00e1rl\u00f3i Szegment\u00e1l\u00e1si Projekt G\u00e9pitanul\u00e1ssal<\/a><\/noindex><\/i>)<\/li>\n
  2. Iris Adatb\u00e1zis<\/a><\/noindex> \u2014 kezd\u0151knek sz\u00f3l\u00f3 adatb\u00e1zis, amely k\u00fcl\u00f6nb\u00f6z\u0151 vir\u00e1gok cs\u00e9szeleveleinek \u00e9s szirmai m\u00e9reteit tartalmazza.<\/li>\n
  3. MNIST Adatb\u00e1zis<\/a><\/noindex> \u2014 k\u00e9zzel \u00edrt sz\u00e1mok adatb\u00e1zisa. 60 000 edz\u0151 k\u00e9p \u00e9s 10 000 teszt k\u00e9p.<\/li>\n
  4. Boston Housing Dataset<\/a><\/noindex> \u2014 n\u00e9pszer\u0171 adatb\u00e1zis mint\u00e1zatfelismer\u00e9shez. Informatika a bostoni h\u00e1zakr\u00f3l: lak\u00e1sok sz\u00e1ma, b\u00e9rleti d\u00edj, b\u0171n\u00f6z\u00e9si index.<\/li>\n
  5. Fake News Detection Dataset<\/a><\/noindex> \u2014 7796 h\u00edrbejegyz\u00e9s jel\u00f6l\u00e9s\u00e9vel: igaz vagy hamis. (Alkalmaz\u00e1si lehet\u0151s\u00e9g Python forr\u00e1ssal: Fake News Detection Python Projekt <\/a><\/noindex><\/i>)<\/li>\n
  6. Bor min\u0151s\u00e9gi adatb\u00e1zis<\/a><\/noindex> \u2014 inform\u00e1ci\u00f3t tartalmaz borokr\u00f3l: 4898 rekord 14 param\u00e9terrel.<\/li>\n
  7. SOCR adatok \u2013 Magass\u00e1g \u00e9s S\u00faly Adatb\u00e1zis<\/a><\/noindex> \u2014 j\u00f3 kiindul\u00f3pont. 25 000 rekordot tartalmaz 18 \u00e9ves emberek magass\u00e1g\u00e1r\u00f3l \u00e9s s\u00faly\u00e1r\u00f3l.
    \n

    \"52<\/a><\/noindex>

    \nA cikket az EDISON Software t\u00e1mogat\u00e1s\u00e1val ford\u00edtott\u00e1k, amely     \u201ekiv\u00e1l\u00f3an\u201d teljes\u00edti a D\u00e9l-K\u00edn\u00e1b\u00f3l \u00e9rkez\u0151 megb\u00edz\u00e1sokat<\/a><\/noindex>, valamint webalkalmaz\u00e1sokat \u00e9s weboldalakat fejleszt<\/a><\/noindex>.<\/p><\/blockquote>\n<\/li>\n

  8. Parkinson Adatb\u00e1zis<\/a><\/noindex> \u2014 195 rekord a Parkinson-k\u00f3rban szenved\u0151 betegekr\u0151l, 25 vizsg\u00e1lati param\u00e9terrel. Haszn\u00e1lhat\u00f3 a betegek \u00e9s eg\u00e9szs\u00e9gesek elk\u00fcl\u00f6n\u00edt\u00e9s\u00e9nek el\u0151zetes \u00e9rt\u00e9kel\u00e9s\u00e9re. (Alkalmaz\u00e1si lehet\u0151s\u00e9g Python forr\u00e1ssal: G\u00e9pitanul\u00e1si Projekt Parkinson-k\u00f3r \u00e9szlel\u00e9s\u00e9re<\/a><\/noindex><\/i>)<\/li>\n
  9. Titanic Adatb\u00e1zis<\/a><\/noindex> \u2014 inform\u00e1ci\u00f3t tartalmaz az utasokr\u00f3l (kor, nem, hozz\u00e1tartoz\u00f3k a fed\u00e9lzeten stb.) 891 a tanul\u00f3 halmazban \u00e9s 418 a teszt halmazban.<\/li>\n
  10. Uber Fuvart\u00f6rt\u00e9netek Adatb\u00e1zis<\/a><\/noindex> \u2014 inform\u00e1ci\u00f3 4,5 milli\u00f3 Uber utaz\u00e1sr\u00f3l 2014-b\u0151l \u00e9s 14 milli\u00f3 2015-b\u0151l. (Alkalmaz\u00e1si lehet\u0151s\u00e9g R forr\u00e1ssal: Uber Adatvizsg\u00e1lati Projekt R-ben<\/a><\/noindex><\/i>)<\/li>\n
  11. Chars74k Adatb\u00e1zis<\/a><\/noindex> \u2014 brit \u00e9s kanadai szimb\u00f3lumok k\u00e9peit tartalmazza 64 oszt\u00e1lyban: 0-9, A-Z, a-z. 7700 7,7k term\u00e9szetes k\u00e9p, 3400k k\u00e9zzel \u00edrt, 62000 sz\u00e1m\u00edt\u00f3g\u00e9p \u00e1ltal szintetiz\u00e1lt bet\u0171st\u00edlus.<\/li>\n
  12. Hitelk\u00e1rtya Csal\u00e1s \u00c9szlel\u00e9si Adatb\u00e1zis<\/a><\/noindex> \u2014 inform\u00e1ci\u00f3t tartalmaz a kompromitt\u00e1lt hitelk\u00e1rty\u00e1k tranzakci\u00f3ir\u00f3l. (Alkalmaz\u00e1si lehet\u0151s\u00e9g forr\u00e1ssal: Hitelk\u00e1rtya Csal\u00e1s \u00c9szlel\u00e9si G\u00e9pitanul\u00e1si Projekt<\/a><\/noindex><\/i>)<\/li>\n
  13. Chatbot Sz\u00e1nd\u00e9kok Adatb\u00e1zis<\/a><\/noindex> \u2014 JSON-f\u00e1jl, amely k\u00fcl\u00f6nb\u00f6z\u0151 tageket tartalmaz: \u00fcdv\u00f6zl\u00e9sek, b\u00facs\u00faz\u00e1s, k\u00f3rh\u00e1zkeres\u00e9s, gy\u00f3gyszert\u00e1rkeres\u00e9s stb. Tartalmaz k\u00e9rd\u00e9s-v\u00e1lasz sablonokat. (Alkalmaz\u00e1si lehet\u0151s\u00e9g Python forr\u00e1ssal: Chatbot Projekt Python-ban<\/a><\/noindex><\/i>)<\/li>\n
  14. Enron Email Adatb\u00e1zis<\/a><\/noindex> \u2014 f\u00e9lmilli\u00f3 e-mailt tartalmaz 150 Enron vezet\u0151t\u0151l.<\/li>\n
  15. A Yelp Adatb\u00e1zis<\/a><\/noindex> \u2014 1,2 milli\u00f3 aj\u00e1nl\u00e1st tartalmaz 1,6 milli\u00f3 felhaszn\u00e1l\u00f3t\u00f3l 1,2 milli\u00f3 szervezetr\u0151l.<\/li>\n
  16. Jeopardy Adatb\u00e1zis<\/a><\/noindex> \u2014 t\u00f6bb mint 200 000 \"k\u00e9rd\u00e9s-v\u00e1lasz\" rekord a n\u00e9pszer\u0171 telev\u00edzi\u00f3s j\u00e1t\u00e9kb\u00f3l.<\/li>\n
  17. Aj\u00e1nl\u00f3 Rendszerek Adatb\u00e1zis<\/a><\/noindex> \u2014 a port\u00e1l az UCSD egyetem adatgy\u0171jtem\u00e9ny\u00e9vel. Tartalmazza a n\u00e9pszer\u0171 weboldalakon (Goodreads, Amazon) tal\u00e1lhat\u00f3 v\u00e9lem\u00e9nyek feljegyz\u00e9seit. Kiv\u00e1l\u00f3an alkalmas aj\u00e1nl\u00f3rendszerek l\u00e9trehoz\u00e1s\u00e1ra. (Alkalmaz\u00e1si lehet\u0151s\u00e9g R forr\u00e1ssal: Filmaj\u00e1nl\u00f3 rendszer projekt R nyelven<\/a><\/noindex> <\/i>)<\/li>\n
  18. UCI Spambase adathalmazon<\/a><\/noindex> \u2014 adathalmaz spamsz\u0171r\u00e9shez. 4601 e-mailt tartalmaz 57 metaadat-param\u00e9terrel.<\/li>\n
  19. Flickr 30k adathalmaz<\/a><\/noindex> \u2014 t\u00f6bb mint 30 000 k\u00e9p \u00e9s azok feliratai. (Flickr 8k adathalmaz<\/a><\/noindex> \u2014 8000 k\u00e9p. Python alap\u00fa projekt: K\u00e9p Felirat Gener\u00e1tor Python Projekt<\/a><\/noindex>)<\/i><\/li>\n
  20. IMDB v\u00e9lem\u00e9nyek<\/a><\/noindex> \u2014 25 000 filmkritika a tanul\u00e1si k\u00e9szletben \u00e9s 25 000 a tesztel\u00e9shez. (Alkalmaz\u00e1si lehet\u0151s\u00e9g R forr\u00e1ssal: Sentiment Analysis Adatb\u00e1ny\u00e1szati Projekt<\/a><\/noindex><\/i>)<\/li>\n
  21. MS COCO adathalmaz<\/a><\/noindex> \u2014 1,5 milli\u00f3 c\u00edmk\u00e9zett k\u00e9p.<\/li>\n
  22. CIFAR-10 \u00e9s CIFAR-100 adathalmaz<\/a><\/noindex> \u2014 a CIFAR-10 60 000 kicsi k\u00e9pet tartalmaz, m\u00e9rete 32*32 pixel, a 0-9 sz\u00e1mjegyekhez. A CIFAR-100 pedig 0-100-ig terjed.<\/li>\n
  23. GTSRB (N\u00e9met k\u00f6zleked\u00e9si t\u00e1bl\u00e1k \u00e9szlel\u00e9si benchmark) adathalmaz<\/a><\/noindex> \u2014 50 000 k\u00e9p 43 k\u00f6zleked\u00e9si t\u00e1bl\u00e1r\u00f3l. (Alkalmaz\u00e1si lehet\u0151s\u00e9g Python forr\u00e1ssal: K\u00f6zleked\u00e9si T\u00e1bl\u00e1k Felismer\u00e9s Python Projekt<\/a><\/noindex><\/i>)<\/li>\n
  24. ImageNet adathalmaz<\/a><\/noindex> \u2014 t\u00f6bb mint 100 000 kifejez\u00e9st \u00e9s k\u00f6r\u00fclbel\u00fcl 1000 k\u00e9pet tartalmaz kifejez\u00e9senk\u00e9nt.<\/li>\n
  25. Mellr\u00e1k Sz\u00f6vettani K\u00e9pek Adathalmaz<\/a><\/noindex> \u2014 az adathalmaz tartalmazza a mellr\u00e1k mint\u00e1it \u00e1br\u00e1zol\u00f3 k\u00e9peket. (Alkalmaz\u00e1si p\u00e9lda Python forr\u00e1ssal a Mellr\u00e1k Oszt\u00e1lyoz\u00f3 Python Projekt<\/a><\/noindex><\/i>)<\/li>\n
  26. Cityscapes adathalmaz<\/a><\/noindex> \u2014 magas min\u0151s\u00e9g\u0171 annot\u00e1ci\u00f3kat tartalmaz k\u00fcl\u00f6nb\u00f6z\u0151 v\u00e1rosok utcai vide\u00f3riportjaiban.<\/li>\n
  27. Kinetics adathalmaz<\/a><\/noindex> \u2014 URL-hivatkoz\u00e1st tartalmaz k\u00f6r\u00fclbel\u00fcl 6,5 milli\u00f3 magas min\u0151s\u00e9g\u0171 videoh\u00f6z.<\/li>\n
  28. MPII emberi p\u00f3z adathalmaz<\/a><\/noindex> \u2014 az adathalmaz 25 000 emberi p\u00f3zt tartalmaz k\u00e9pekkel \u00e9s \u00edz\u00fcleti annot\u00e1ci\u00f3val.<\/li>\n
  29. 20BN-something-something adathalmaz v2<\/a><\/noindex> \u2014 magas min\u0151s\u00e9g\u0171 vide\u00f3k gy\u0171jtem\u00e9nye, amelyek bemutatj\u00e1k, hogyan v\u00e9grehajt valaki k\u00fcl\u00f6nf\u00e9le cselekv\u00e9seket.<\/li>\n
  30. Objektum 365 adathalmaz<\/a><\/noindex> \u2014 az adathalmaz magas min\u0151s\u00e9g\u0171 k\u00e9peket tartalmaz, amelyekben objektumok keretezve vannak.<\/li>\n
  31. Fot\u00f3rajz adathalmaz<\/a><\/noindex> \u2014 t\u00f6bb mint 1000 k\u00e9pet tartalmaz a kont\u00farv\u00e1zlataikkal.<\/li>\n
  32. CQ500 adathalmaz<\/a><\/noindex> \u2014 az adathalmaz 491 fej CT-vizsg\u00e1latot tartalmaz 193 317 szelettel.<\/li>\n
  33. IMDB-Wiki adathalmaz<\/a><\/noindex> \u2014 az adathalmaz t\u00f6bb mint 5 milli\u00f3 arck\u00e9pet tartalmaz, nem \u00e9s kor megjel\u00f6l\u00e9ssel. (Alkalmaz\u00e1si p\u00e9lda Python forr\u00e1ssal a Nem & Kor Meghat\u00e1roz\u00f3 Python Projekt<\/a><\/noindex><\/i>)<\/li>\n
  34. Youtube 8M adathalmaz<\/a><\/noindex> \u2014 c\u00edmk\u00e9zett videoadathalmaz, amely 6,1 milli\u00f3 Youtube vide\u00f3azonos\u00edt\u00f3t tartalmaz.<\/li>\n
  35. Urban Sound 8K adathalmaz<\/a><\/noindex> \u2014 v\u00e1rosi hangok adathalmaz (8732 v\u00e1rosi hangot tartalmaz 10 oszt\u00e1lyb\u00f3l).<\/li>\n
  36. LSUN adathalmaz<\/a><\/noindex> \u2014 adathalmaz, amely milli\u00f3nyi sz\u00ednes k\u00e9pet tartalmaz jelenetekr\u0151l \u00e9s objektumokr\u00f3l (kb. 59 milli\u00f3 k\u00e9p, 10 k\u00fcl\u00f6nb\u00f6z\u0151 jelenetkateg\u00f3ri\u00e1val \u00e9s 20 k\u00fcl\u00f6nb\u00f6z\u0151 objektumkateg\u00f3ri\u00e1val).<\/li>\n
  37. RAVDESS adathalmaz<\/a><\/noindex> \u2014 audiovizu\u00e1lis adatb\u00e1zis \u00e9rzelmi besz\u00e9dr\u0151l. (Alkalmaz\u00e1si p\u00e9lda Python forr\u00e1ssal a Besz\u00e9d \u00c9rzelem Felismer\u0151 Python Projekt<\/a><\/noindex><\/i>)<\/li>\n
  38. Librispeech adathalmaz<\/a><\/noindex> \u2014 a dataset that contains 1000 hours of English speech with various accents.<\/li>\n
  39. Baidu Apolloscape Dataset<\/a><\/noindex> \u2014 a dataset for the development of autonomous driving technologies. <\/li>\n
  40. Quandl Data Portal<\/a><\/noindex> \u2014 a repository of economic and financial data (both free and paid content available).<\/li>\n
  41. The World Bank Open Data Portal<\/a><\/noindex> \u2014 information on loans issued by the World Bank to developing countries.<\/li>\n
  42. IMF Data Portal<\/a><\/noindex> \u2014 the portal of the International Monetary Fund that publishes data on international finance, debt rates, investments, currency reserves, and commodities.<\/li>\n
  43. American Economic Association (AEA) Data Portal<\/a><\/noindex> \u2014 a resource for finding macroeconomic data of the USA.<\/li>\n
  44. Google Trends Data Portal<\/a><\/noindex> \u2014 data on Google trends can be used for visual examination and analysis of data.<\/li>\n
  45. Financial Times Market Data Portal<\/a><\/noindex> \u2014 a resource for obtaining up-to-date information on financial markets from around the world.<\/li>\n
  46. Data.gov Portal<\/a><\/noindex> \u2014 the open data portal of the US government (agriculture, health, climate, education, energy, finance, science and research, etc.).<\/li>\n
  47. Data Portal: Open government data (India)<\/a><\/noindex> \u2014 the open government data platform of India.<\/li>\n
  48. Food environment Atlas Data Portal<\/a><\/noindex> \u2014 contains research data on nutrition in the USA.<\/li>\n
  49. Health Data Portal<\/a><\/noindex> \u2014 this is the portal of the US Department of Health and Human Services.<\/li>\n
  50. Centers for Disease Control and Prevention Data Portal<\/a><\/noindex> \u2014 contains a wide range of health-related data.<\/li>\n
  51. London Datastore Portal<\/a><\/noindex> \u2014 data about the lives of people in London.<\/li>\n
  52. Canada Government Open Data Portal<\/a><\/noindex> \u2014 the open data portal about Canadians (agriculture, arts, music, education, government, health, etc.)<\/li>\n<\/ol>\n

    Tov\u00e1bbi olvas\u00e1s<\/h3>\n

    <\/p>\n