{"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\/ro\/blog\/news\/9-podhodov-dlya-vyyavleniya-anomalij","title":{"rendered":"9 abord\u0103ri pentru identificarea anomaliilor","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

\u00cen articolul anterior<\/a><\/noindex> Am discutat despre prognozarea seriei temporale. O continuare logic\u0103 ar fi un articol despre detectarea anomaliilor. <\/p>\n

Aplicare<\/h3>\n

\nDetectarea anomaliilor este utilizat\u0103 \u00een domenii precum: <\/p>\n

1) Predic\u021bia defect\u0103rii echipamentului<\/h3>\n

\nAstfel, \u00een 2010, centrifugele din Iran au fost atacate de virusul Stuxnet, care a impus un mod de func\u021bionare neoptimizat \u0219i a deteriorat o parte din echipament din cauza uzurii accelerate. <\/p>\n

Dac\u0103 pe echipament s-ar fi folosit algoritmi de c\u0103utare a anomaliilor, situa\u021biile de defectare ar fi putut fi evitate. <\/p>\n

\"9
\n<\/a><\/noindex>
\nC\u0103utarea anomaliilor \u00een func\u021bionarea echipamentului nu este utilizat\u0103 doar \u00een industria nuclear\u0103, ci \u0219i \u00een metalurgie \u0219i \u00een func\u021bionarea turbinelor aviatice. \u0218i \u00een alte domenii unde utilizarea diagnostic\u0103rii predictive este mai ieftin\u0103 dec\u00e2t posibilele pierderi cauzate de defect\u0103ri imprevizibile.<\/p>\n

2) Predic\u021bia fraudelor<\/h3>\n

\nDac\u0103 bani sunt retr\u0103\u0219i de pe cardul pe care \u00eel folose\u0219ti \u00een Podolsk, \u00een Albania, este posibil ca tranzac\u021biile s\u0103 merite o verificare suplimentar\u0103.<\/p>\n

3) Detectarea modelelor de consum anormale<\/h3>\n

\nDac\u0103 o parte din clien\u021bi demonstreaz\u0103 un comportament anormal, este posibil s\u0103 existe o problem\u0103 despre care nu \u0219tii.<\/p>\n

4) Detectarea cererii \u0219i \u00eenc\u0103rc\u0103rii anormale<\/h3>\n

\nDac\u0103 v\u00e2nz\u0103rile dintr-un magazin FMCG au sc\u0103zut sub limita intervalului de \u00eencredere al prognozei, merit\u0103 s\u0103 g\u0103se\u0219ti cauza acestui fenomen.<\/p>\n

Aproape de detectarea anomaliilor<\/h2>\n

<\/p>\n

1) Metoda vectorilor de suport cu un singur clasa One-Class SVM<\/h3>\n

\nSe potrive\u0219te atunci c\u00e2nd datele din setul de antrenament se supun unei distrib\u021bii normale, iar cele din test con\u021bin anomalii.<\/p>\n

Metoda cu un singur clasa a vectorilor de suport construie\u0219te o suprafa\u021b\u0103 neliniar\u0103 \u00een jurul originii coordonatelor. Este posibil s\u0103 se stabileasc\u0103 o limit\u0103 de t\u0103iere pentru a determina ce date sunt considerate anormale.<\/p>\n

Din experien\u021ba echipei noastre DATA4, One-Class SVM este cel mai frecvent algoritm utilizat pentru a aborda problema c\u0103ut\u0103rii anomaliilor. <\/p>\n

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

2) Metoda p\u0103durii de izolare \u2013 isolate forest<\/h3>\n

\nPrin metoda de construc\u021bie aleatorie a copacilor, valorile anormale vor ajunge \u00een frunze \u00een etapele ini\u021biale (la o ad\u00e2ncime mic\u0103 a copacului), adic\u0103 valorile anormale sunt mai u\u0219or de \u201eizolat\u201d. Eviden\u021bierea valorilor anormale se face \u00een primele itera\u021bii ale func\u021bion\u0103rii algoritmului.<\/p>\n

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

3) \u00cenveli\u0219ul eliptic \u0219i metodele statistice<\/h3>\n

\nEste utilizat atunci c\u00e2nd datele sunt distribuite normal. Cu c\u00e2t o m\u0103surare este mai aproape de coada mixturii de distribu\u021bii, cu at\u00e2t valoarea este mai anormal\u0103.<\/p>\n

Aceast\u0103 clas\u0103 include \u0219i alte metode statistice.<\/p>\n

\"9
\n
\n\"9
\nImagine de pe site-ul dyakonov.org<\/i><\/p>\n

4) Metodele metrice<\/h3>\n

\nPrintre metode se num\u0103r\u0103 algoritmii, cum ar fi k-nearest neighbors, k-vecinul cel mai apropiat, ABOD (detectarea anomaliilor bazat\u0103 pe unghi) sau LOF (factorul de anomalii local).<\/p>\n

Acestea sunt potrivite dac\u0103 distan\u021ba dintre valorile caracteristicilor este egal\u0103 sau normalizat\u0103 (pentru a nu m\u0103sura \u0219arpele \u00een papagali). <\/p>\n

Algoritmul k-nearest neighbors presupune c\u0103 valorile normale sunt situate \u00eentr-o anumit\u0103 zon\u0103 a spa\u021biului multidimensional, iar distan\u021ba fa\u021b\u0103 de anomalii va fi mai mare dec\u00e2t fa\u021b\u0103 de hiperplanul de separare. <\/p>\n

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

5) Metodele de clustering<\/h3>\n

\nEsenta metodelor de clustering este c\u0103 dac\u0103 o valoare este \u00eendep\u0103rtat\u0103 de centrele cluster-elor cu mai mult dec\u00e2t o anumit\u0103 m\u0103rime, valoarea poate fi considerat\u0103 anormal\u0103. <\/p>\n

Principalul lucru este s\u0103 folose\u0219ti un algoritm care clusterizeaz\u0103 corect datele, ceea ce depinde de sarcina specific\u0103. <\/p>\n

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

6) Metoda componentelor principale<\/h3>\n

\nEste potrivit\u0103 acolo unde se eviden\u021biaz\u0103 direc\u021biile de cea mai mare schimbare a dispersiei.<\/p>\n

7) Algoritmi bazati pe prognozarea seriilor temporale<\/h3>\n

\nIdeea este c\u0103 dac\u0103 o valoare deviaz\u0103 din intervalul de \u00eencredere al prognozei, valoarea este considerat\u0103 anormal\u0103. Pentru prognozarea seriilor temporale se folosesc algoritmi precum netezirea tripl\u0103, S(ARIMA), boosting etc.<\/p>\n

Despre algoritmii de prognozare a seriilor temporale s-a vorbit \u00een articolul precedent.<\/p>\n

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

8) \u00cenv\u0103\u021bare supravegheat\u0103 (regresie, clasificare)<\/h3>\n

\nDac\u0103 datele permit, folosim algoritmi de la regresia liniar\u0103 la re\u021bele neuronale recurente. Vom m\u0103sura diferen\u021ba dintre prognoz\u0103 \u0219i valoarea real\u0103 \u0219i vom trasa concluzia despre c\u00e2t de mult datele deviaz\u0103 de la normal. Este important ca algoritmul s\u0103 aib\u0103 o capacitate de generalizare suficient\u0103 \u0219i ca setul de antrenament s\u0103 nu con\u021bin\u0103 valori anormale.<\/p>\n

9) Teste de model<\/h3>\n

\nVom aborda problema identific\u0103rii anomaliilor ca pe o problem\u0103 de generare de recomand\u0103ri. Vom descompune matricea noastr\u0103 de caracteristici folosind SVD sau ma\u0219ini de factorizare \u0219i valorile din noua matrice, care se abate semnificativ de la cele originale, le vom considera anormale. <\/p>\n

\"9
\n
\nImagine de pe site-ul dyakonov.org<\/i><\/p>\n

Concluzie<\/h3>\n

\n\u00cen acest articol am analizat principalele abord\u0103ri pentru detectarea anomaliilor. <\/p>\n

C\u0103utarea anomaliilor poate fi considerat\u0103, \u00eentr-o mare m\u0103sur\u0103, o art\u0103. Nu exist\u0103 un algoritm sau o metod\u0103 perfect\u0103 care s\u0103 rezolve toate problemele. De regul\u0103, se folose\u0219te un set de metode pentru a aborda un caz specific. C\u0103utarea anomaliilor se realizeaz\u0103 prin intermediul metodei SVM (Support Vector Machine), p\u0103durilor izolate, metode metrice \u0219i de clustering, precum \u0219i prin utilizarea componentelor principale \u0219i predic\u021bia seriilor temporale.<\/p>\n

Dac\u0103 \u0219ti\u021bi alte metode, scrie\u021bi despre ele \u00een comentarii la articol.
\n
Sursa: 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\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 abord\u0103ri pentru identificarea anomaliilor | ProHoster","description":"\u00cen articolul anterior, am discutat despre predic\u021bia seriilor temporale. O continuitate logic\u0103 ar fi s\u0103 abord\u0103m subiectul identific\u0103rii anomaliilor.","canonical_url":"https:\/\/prohoster.info\/ro\/blog\/news\/9-podhodov-dlya-vyyavleniya-anomalij","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"ro_RO","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\/ro\/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\/ro\/wp-json\/wp\/v2\/posts\/53278","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/ro\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/ro\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/ro\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/ro\/wp-json\/wp\/v2\/comments?post=53278"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/ro\/wp-json\/wp\/v2\/posts\/53278\/revisions"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/ro\/wp-json\/wp\/v2\/media?parent=53278"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/ro\/wp-json\/wp\/v2\/categories?post=53278"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/ro\/wp-json\/wp\/v2\/tags?post=53278"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}