{"id":53278,"date":"2019-11-28T00:00:00","date_gmt":"2019-11-27T21:00:00","guid":{"rendered":"https:\/\/prohoster.info\/blog\/blog_prohoster\/9-podhodov-dlya-vyyavleniya-anomalij"},"modified":"2020-02-18T14:01:10","modified_gmt":"2020-02-18T11:01:10","slug":"9-podhodov-dlya-vyyavleniya-anomalij","status":"publish","type":"post","link":"https:\/\/prohoster.info\/cs\/blog\/news\/9-podhodov-dlya-vyyavleniya-anomalij","title":{"rendered":"9 p\u0159\u00edstup\u016f k odhalov\u00e1n\u00ed anom\u00e1li\u00ed","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

V p\u0159edchoz\u00ed \u010dl\u00e1nku<\/a><\/noindex> Hovo\u0159ili jsme o predikci \u010dasov\u00fdch \u0159ad. Logick\u00fdm pokra\u010dov\u00e1n\u00edm bude \u010dl\u00e1nek o detekci anom\u00e1li\u00ed. <\/p>\n

Pou\u017eit\u00ed<\/h3>\n

\nDetekce anom\u00e1li\u00ed se vyu\u017e\u00edv\u00e1 v oblastech jako jsou: <\/p>\n

1) Predikce poruch za\u0159\u00edzen\u00ed<\/h3>\n

\nV roce 2010 byly \u00edr\u00e1nsk\u00e9 centrifugy napadeny virem Stuxnet, kter\u00fd nastavil suboptim\u00e1ln\u00ed provozn\u00ed re\u017eim a zni\u010dil \u010d\u00e1st za\u0159\u00edzen\u00ed v d\u016fsledku urychlen\u00e9ho opot\u0159eben\u00ed. <\/p>\n

Kdyby na za\u0159\u00edzen\u00ed byly pou\u017eity algoritmy pro detekci anom\u00e1li\u00ed, t\u00e9to situaci by se dalo p\u0159edej\u00edt. <\/p>\n

\"9
\n<\/a><\/noindex>
\nDetekce anom\u00e1li\u00ed v provozu za\u0159\u00edzen\u00ed se vyu\u017e\u00edv\u00e1 nejen v jadern\u00e9m pr\u016fmyslu, ale tak\u00e9 v metalurgii a v oblasti leteck\u00fdch turb\u00edn. A v dal\u0161\u00edch oblastech, kde vyu\u017eit\u00ed prediktivn\u00ed diagnostiky je levn\u011bj\u0161\u00ed ne\u017e mo\u017en\u00e9 ztr\u00e1ty p\u0159i nep\u0159edv\u00eddateln\u00fdch poruch\u00e1ch.<\/p>\n

2) Predikce podvodn\u00fdch \u010dinnost\u00ed<\/h3>\n

\nPokud se z karty, kterou pou\u017e\u00edv\u00e1te v Podolsku, vyb\u00edraj\u00ed pen\u00edze v Alb\u00e1nii, mo\u017en\u00e1 by m\u011bly b\u00fdt transakce dodate\u010dn\u011b prov\u011b\u0159eny.<\/p>\n

3) Detekce anom\u00e1ln\u00edch spot\u0159ebitelsk\u00fdch vzorc\u016f<\/h3>\n

\nPokud \u010d\u00e1st z\u00e1kazn\u00edk\u016f vykazuje anom\u00e1ln\u00ed chov\u00e1n\u00ed, mo\u017en\u00e1 je probl\u00e9m, o kter\u00e9m nev\u00edte.<\/p>\n

4) Detekce anom\u00e1ln\u00ed popt\u00e1vky a zat\u00ed\u017een\u00ed<\/h3>\n

\nPokud prodeje v obchod\u011b FMCG klesly pod mez d\u016fv\u011bryhodn\u00e9ho intervalu predikce, je t\u0159eba zjistit p\u0159\u00ed\u010dinu tohoto d\u011bn\u00ed.<\/p>\n

P\u0159\u00edstupy k detekci anom\u00e1li\u00ed<\/h2>\n

<\/p>\n

1) Metoda podp\u016frn\u00fdch vektor\u016f s jednou t\u0159\u00eddou One-Class SVM<\/h3>\n

\nVhodn\u00e1, kdy\u017e v tr\u00e9ninkov\u00e9m souboru data podl\u00e9haj\u00ed norm\u00e1ln\u00edmu rozd\u011blen\u00ed, zat\u00edmco v testovac\u00edm souboru obsahuj\u00ed anom\u00e1lie.<\/p>\n

Jednoduch\u00e1 metoda podp\u016frn\u00fdch vektor\u016f stav\u00ed neline\u00e1rn\u00ed plochu kolem po\u010d\u00e1tku sou\u0159adnic. Je mo\u017en\u00e9 nastavit prahovou hodnotu pro to, jak\u00e9 \u00fadaje pova\u017eovat za anom\u00e1ln\u00ed.<\/p>\n

Na z\u00e1klad\u011b zku\u0161enost\u00ed na\u0161eho t\u00fdmu DATA4 je One-Class SVM nej\u010dast\u011bji pou\u017e\u00edvan\u00fd algoritmus pro \u0159e\u0161en\u00ed \u00falohy detekce anom\u00e1li\u00ed. <\/p>\n

\"9
\n<\/p>\n

2) Metoda izola\u010dn\u00edho lesa \u2013 isolate forest<\/h3>\n

\nP\u0159i \u201en\u00e1hodn\u00e9m\u201c zp\u016fsobu stavby strom\u016f budou odlehl\u00e9 hodnoty spadat do list\u016f na ran\u00fdch f\u00e1z\u00edch (na mal\u00e9 hloubce stromu), tedy bude jednodu\u0161\u0161\u00ed tyto odlehl\u00e9 hodnoty \"izolovat\". Odd\u011blen\u00ed anom\u00e1ln\u00edch hodnot prob\u00edh\u00e1 v prvn\u00edch iterac\u00edch pr\u00e1ce algoritmu.<\/p>\n

\"9
\n<\/p>\n

3) Elliptic envelope a statistick\u00e9 metody<\/h3>\n

\nPou\u017e\u00edv\u00e1 se, kdy\u017e jsou data norm\u00e1ln\u011b rozd\u011blena. \u010c\u00edm bl\u00ed\u017ee je m\u011b\u0159en\u00ed k ocasu sm\u011bsi rozd\u011blen\u00ed, t\u00edm v\u00edce je hodnota anom\u00e1ln\u00ed.<\/p>\n

Tato t\u0159\u00edda zahrnuje i dal\u0161\u00ed statistick\u00e9 metody.<\/p>\n

\"9
\n
\n\"9
\nObr\u00e1zek z webu dyakonov.org<\/i><\/p>\n

4) Metodick\u00e9 metody<\/h3>\n

\nMezi metody pat\u0159\u00ed algoritmy, jako jsou k nejbli\u017e\u0161\u00edch soused\u016f, k-t\u00fd nejbli\u017e\u0161\u00ed soused, ABOD (detekce odlehl\u00fdch bod\u016f na z\u00e1klad\u011b \u00fahlu) nebo LOF (m\u00edstn\u00ed faktor odlehlosti).<\/p>\n

Jsou vhodn\u00e9, pokud je vzd\u00e1lenost mezi hodnotami atribut\u016f ekvivalentn\u00ed nebo normalizov\u00e1na (abychom nem\u011b\u0159ili hady v papou\u0161c\u00edch). <\/p>\n

Algoritmus k nejbli\u017e\u0161\u00edch soused\u016f p\u0159edpokl\u00e1d\u00e1, \u017ee norm\u00e1ln\u00ed hodnoty jsou um\u00edst\u011bny v ur\u010dit\u00e9 oblasti v\u00edcerozm\u011brn\u00e9ho prostoru, zat\u00edmco vzd\u00e1lenost k anom\u00e1li\u00edm bude v\u011bt\u0161\u00ed ne\u017e k d\u011bl\u00edc\u00ed hyperrovnin\u011b. <\/p>\n

\"9
\n<\/p>\n

5) Klastrick\u00e9 metody<\/h3>\n

\nPodstatou klastrick\u00fdch metod je, \u017ee pokud je hodnota vzd\u00e1len\u00e1 od center skupin v\u00edce ne\u017e ur\u010ditou m\u00edru, m\u016f\u017ee b\u00fdt pova\u017eov\u00e1na za anom\u00e1ln\u00ed. <\/p>\n

Hlavn\u00ed je pou\u017e\u00edvat algoritmus, kter\u00fd spr\u00e1vn\u011b klastruje data, co\u017e z\u00e1vis\u00ed na konkr\u00e9tn\u00ed \u00faloze. <\/p>\n

\"9
\n<\/p>\n

6) Metoda hlavn\u00edch komponent<\/h3>\n

\nVhodn\u00e9, kde se zd\u016fraz\u0148uj\u00ed sm\u011bry s nejv\u011bt\u0161\u00ed zm\u011bnou rozptylu.<\/p>\n

7) Algoritmy na z\u00e1klad\u011b predikce \u010dasov\u00fdch \u0159ad<\/h3>\n

\nMy\u0161lenka spo\u010d\u00edv\u00e1 v tom, \u017ee pokud hodnota vybo\u010duje z intervalov\u00e9ho odhadu, pova\u017euje se za anom\u00e1ln\u00ed. Pro predikci \u010dasov\u00e9 \u0159ady se pou\u017e\u00edvaj\u00ed algoritmy jako trojit\u00e9 vyhlazov\u00e1n\u00ed, S(ARIMA), boosting atd.<\/p>\n

O algoritmech pro predikci \u010dasov\u00fdch \u0159ad se hovo\u0159ilo v p\u0159edchoz\u00edm \u010dl\u00e1nku.<\/p>\n

\"9
\n<\/p>\n

8) U\u010den\u00ed se s u\u010ditelem (regrese, klasifikace)<\/h3>\n

\nPokud to data dovol\u00ed, pou\u017e\u00edv\u00e1me algoritmy od line\u00e1rn\u00ed regrese po rekurentn\u00ed s\u00edt\u011b. Zm\u011b\u0159\u00edme rozd\u00edl mezi odhadem a skute\u010dnou hodnotou a ur\u010d\u00edme, jak moc data vybo\u010duj\u00ed z normy. D\u016fle\u017eit\u00e9 je, aby algoritmus m\u011bl dostate\u010dn\u00e9 generaliza\u010dn\u00ed schopnosti a tr\u00e9ninkov\u00e1 data neobsahovala anom\u00e1ln\u00ed hodnoty.<\/p>\n

9) Modelov\u00e9 testy<\/h3>\n

\nP\u0159istoup\u00edme k \u00faloze detekce anom\u00e1li\u00ed, jako bychom se zab\u00fdvali doporu\u010dovac\u00edmi syst\u00e9my. Rozlo\u017e\u00edme na\u0161i matici atribut\u016f pomoc\u00ed SVD nebo faktoriza\u010dn\u00edch stroj\u016f a hodnoty v nov\u00e9 matici, kter\u00e9 se podstatn\u011b li\u0161\u00ed od p\u016fvodn\u00edch, pova\u017eujeme za anom\u00e1ln\u00ed. <\/p>\n

\"9
\n
\nObr\u00e1zek z webu dyakonov.org<\/i><\/p>\n

Z\u00e1v\u011br<\/h3>\n

\nV tomto \u010dl\u00e1nku jsme prozkoumali hlavn\u00ed p\u0159\u00edstupy k detekci anom\u00e1li\u00ed. <\/p>\n

Vyhled\u00e1v\u00e1n\u00ed anom\u00e1li\u00ed lze do zna\u010dn\u00e9 m\u00edry nazvat um\u011bn\u00edm. Neexistuje ide\u00e1ln\u00ed algoritmus nebo p\u0159\u00edstup, jeho\u017e pou\u017eit\u00ed by vy\u0159e\u0161ilo v\u0161echny \u00fakoly. \u010cast\u011bji se pou\u017e\u00edv\u00e1 komplex metod pro \u0159e\u0161en\u00ed konkr\u00e9tn\u00edho p\u0159\u00edpadu. Vyhled\u00e1v\u00e1n\u00ed anom\u00e1li\u00ed se prov\u00e1d\u00ed pomoc\u00ed jednovrstv\u00e9ho metody podp\u016frn\u00fdch vektor\u016f, izola\u010dn\u00edch les\u016f, metrick\u00fdch a klastrov\u00fdch metod, stejn\u011b jako s vyu\u017eit\u00edm hlavn\u00edch komponent a predikce \u010dasov\u00fdch \u0159ad.<\/p>\n

Pokud zn\u00e1te dal\u0161\u00ed metody, napi\u0161te o nich do koment\u00e1\u0159\u016f k \u010dl\u00e1nku.
\n
Zdroj: habr.com<\/a><\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"

\u0412 \u043f\u0440\u0435\u0434\u044b\u0434\u0443\u0449\u0435\u0439 \u0441\u0442\u0430\u0442\u044c\u0435 \u043c\u044b \u0433\u043e\u0432\u043e\u0440\u0438\u043b\u0438 \u043e \u043f\u0440\u043e\u0433\u043d\u043e\u0437\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u0438 \u0432\u0440\u0435\u043c\u0435\u043d\u043d\u044b\u0445 \u0440\u044f\u0434\u043e\u0432. \u041b\u043e\u0433\u0438\u0447\u043d\u044b\u043c \u043f\u0440\u043e\u0434\u043e\u043b\u0436\u0435\u043d\u0438\u0435\u043c \u0431\u0443\u0434\u0435\u0442 \u0441\u0442\u0430\u0442\u044c\u044f \u043e \u0432\u044b\u044f\u0432\u043b\u0435\u043d\u0438\u0438 \u0430\u043d\u043e\u043c\u0430\u043b\u0438\u0439. \u041f\u0440\u0438\u043c\u0435\u043d\u0435\u043d\u0438\u0435 \u0412\u044b\u044f\u0432\u043b\u0435\u043d\u0438\u0435 \u0430\u043d\u043e\u043c\u0430\u043b\u0438\u0439 \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u0443\u0435\u0442\u0441\u044f \u0432 \u0442\u0430\u043a\u0438\u0445 \u043e\u0431\u043b\u0430\u0441\u0442\u044f\u0445 \u043a\u0430\u043a: 1) \u041f\u0440\u0435\u0434\u0441\u043a\u0430\u0437\u0430\u043d\u0438\u0435 \u043f\u043e\u043b\u043e\u043c\u043e\u043a \u043e\u0431\u043e\u0440\u0443\u0434\u043e\u0432\u0430\u043d\u0438\u044f \u0422\u0430\u043a, \u0432 2010 \u0433\u043e\u0434\u0443 \u0418\u0440\u0430\u043d\u0441\u043a\u0438\u0435 \u0446\u0435\u043d\u0442\u0440\u0438\u0444\u0443\u0433\u0438 \u0431\u044b\u043b\u0438 \u0430\u0442\u0430\u043a\u043e\u0432\u0430\u043d\u044b \u0432\u0438\u0440\u0443\u0441\u043e\u043c Stuxnet, \u043a\u043e\u0442\u043e\u0440\u044b\u0439 \u0437\u0430\u0434\u0430\u043b \u043d\u0435\u043e\u043f\u0442\u0438\u043c\u0430\u043b\u044c\u043d\u044b\u0439 \u0440\u0435\u0436\u0438\u043c \u0440\u0430\u0431\u043e\u0442\u044b \u043e\u0431\u043e\u0440\u0443\u0434\u043e\u0432\u0430\u043d\u0438\u044f \u0438 \u0432\u044b\u0432\u0435\u043b \u0438\u0437 \u0441\u0442\u0440\u043e\u044f \u0447\u0430\u0441\u0442\u044c \u043e\u0431\u043e\u0440\u0443\u0434\u043e\u0432\u0430\u043d\u0438\u044f \u0437\u0430 \u0441\u0447\u0435\u0442 \u0443\u0441\u043a\u043e\u0440\u0435\u043d\u043d\u043e\u0433\u043e \u0438\u0437\u043d\u043e\u0441\u0430. \u0415\u0441\u043b\u0438 […]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[702],"tags":[],"class_list":["post-53278","post","type-post","status-publish","format-standard","hentry","category-news"],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t\n\t\n\t\n\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\n","aioseo_head_json":{"title":"\ud83e\udd479 p\u0159\u00edstup\u016f k odhalov\u00e1n\u00ed anom\u00e1li\u00ed | ProHoster","description":"V p\u0159edchoz\u00edm \u010dl\u00e1nku jsme hovo\u0159ili o predikci \u010dasov\u00fdch \u0159ad. Logick\u00fdm pokra\u010dov\u00e1n\u00edm bude \u010dl\u00e1nek o odhalov\u00e1n\u00ed anom\u00e1li\u00ed.","canonical_url":"https:\/\/prohoster.info\/cs\/blog\/news\/9-podhodov-dlya-vyyavleniya-anomalij","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"cs_CZ","og:site_name":"ProHoster | \u041a\u0443\u043f\u0438\u0442\u044c \u043d\u0430\u0434\u0435\u0436\u043d\u044b\u0439 \u0445\u043e\u0441\u0442\u0438\u043d\u0433 \u0434\u043b\u044f \u0441\u0430\u0439\u0442\u043e\u0432 \u0441 \u0437\u0430\u0449\u0438\u0442\u043e\u0439 \u043e\u0442 DDoS, VPS VDS \u0441\u0435\u0440\u0432\u0435\u0440\u044b","og:type":"article","og:title":"\ud83e\udd479 \u043f\u043e\u0434\u0445\u043e\u0434\u043e\u0432 \u0434\u043b\u044f \u0432\u044b\u044f\u0432\u043b\u0435\u043d\u0438\u044f \u0430\u043d\u043e\u043c\u0430\u043b\u0438\u0439 | ProHoster","og:description":"\u0412 \u043f\u0440\u0435\u0434\u044b\u0434\u0443\u0449\u0435\u0439 \u0441\u0442\u0430\u0442\u044c\u0435 \u043c\u044b \u0433\u043e\u0432\u043e\u0440\u0438\u043b\u0438 \u043e \u043f\u0440\u043e\u0433\u043d\u043e\u0437\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u0438 \u0432\u0440\u0435\u043c\u0435\u043d\u043d\u044b\u0445 \u0440\u044f\u0434\u043e\u0432. \u041b\u043e\u0433\u0438\u0447\u043d\u044b\u043c \u043f\u0440\u043e\u0434\u043e\u043b\u0436\u0435\u043d\u0438\u0435\u043c \u0431\u0443\u0434\u0435\u0442 \u0441\u0442\u0430\u0442\u044c\u044f \u043e \u0432\u044b\u044f\u0432\u043b\u0435\u043d\u0438\u0438 \u0430\u043d\u043e\u043c\u0430\u043b\u0438\u0439.","og:url":"https:\/\/prohoster.info\/cs\/blog\/news\/9-podhodov-dlya-vyyavleniya-anomalij","og:image":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:secure_url":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:width":350,"og:image:height":350,"article:published_time":"2019-11-27T21:00:00+00:00","article:modified_time":"2020-02-18T11:01:10+00:00","article:publisher":"https:\/\/www.facebook.com\/prohoster","article:author":"https:\/\/www.facebook.com\/prohoster"},"aioseo_meta_data":{"post_id":"53278","title":null,"description":null,"keywords":null,"keyphrases":null,"primary_term":null,"canonical_url":null,"og_title":null,"og_description":null,"og_object_type":"default","og_image_type":"default","og_image_url":null,"og_image_width":null,"og_image_height":null,"og_image_custom_url":null,"og_image_custom_fields":null,"og_video":null,"og_custom_url":null,"og_article_section":null,"og_article_tags":null,"twitter_use_og":false,"twitter_card":"default","twitter_image_type":"default","twitter_image_url":null,"twitter_image_custom_url":null,"twitter_image_custom_fields":null,"twitter_title":null,"twitter_description":null,"schema":{"blockGraphs":[],"customGraphs":[],"default":{"data":{"Article":[],"Course":[],"Dataset":[],"FAQPage":[],"Movie":[],"Person":[],"Product":[],"ProductReview":[],"Car":[],"Recipe":[],"Service":[],"SoftwareApplication":[],"WebPage":[]},"graphName":"","isEnabled":true},"graphs":[]},"schema_type":null,"schema_type_options":null,"pillar_content":false,"robots_default":true,"robots_noindex":false,"robots_noarchive":false,"robots_nosnippet":false,"robots_nofollow":false,"robots_noimageindex":false,"robots_noodp":false,"robots_notranslate":false,"robots_max_snippet":null,"robots_max_videopreview":null,"robots_max_imagepreview":"large","priority":null,"frequency":null,"local_seo":null,"seo_analyzer_scan_date":"2026-01-24 06:46:20","breadcrumb_settings":null,"limit_modified_date":false,"reviewed_by":null,"ai":null,"created":"2021-02-28 20:28:47","updated":"2026-01-24 06:46:20","focus_keyword":null,"additional_keywords":null,"truseo_locale":null},"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/posts\/53278","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/comments?post=53278"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/posts\/53278\/revisions"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/media?parent=53278"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/categories?post=53278"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/tags?post=53278"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}