{"id":39546,"date":"2019-10-31T22:35:27","date_gmt":"2019-10-31T19:35:27","guid":{"rendered":"https:\/\/prohoster.info\/blog\/kak-my-ispolzuem-tsepi-markova-v-otsenke-reshenij-i-poiske-bagov-so-skriptom-na-python\/"},"modified":"2019-10-31T22:35:27","modified_gmt":"2019-10-31T19:35:27","slug":"kak-my-ispolzuem-tsepi-markova-v-otsenke-reshenij-i-poiske-bagov-so-skriptom-na-python","status":"publish","type":"post","link":"https:\/\/prohoster.info\/sq\/blog\/news\/kak-my-ispolzuem-tsepi-markova-v-otsenke-reshenij-i-poiske-bagov-so-skriptom-na-python","title":{"rendered":"Si p\u00ebrdorim zinxhir\u00ebt Markov n\u00eb vler\u00ebsimin e vendimeve dhe k\u00ebrkimin e gabimeve. Me skript n\u00eb Python","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

Na \u00ebsht\u00eb e r\u00ebnd\u00ebsishme t\u00eb kuptojm\u00eb se \u00e7far\u00eb ndodh me student\u00ebt tan\u00eb gjat\u00eb m\u00ebsimit dhe si k\u00ebto ngjarje ndikojn\u00eb n\u00eb rezultat, prandaj ne krijojm\u00eb Customer Journey Map \u2014 hart\u00ebn e p\u00ebrvoj\u00ebs s\u00eb klientit. Sepse procesi i m\u00ebsimit nuk \u00ebsht\u00eb nj\u00eb di\u00e7ka e vazhdueshme dhe e t\u00ebr\u00eb, \u00ebsht\u00eb nj\u00eb zinxhir ngjarjesh dhe veprimesh t\u00eb lidhura me studentin, ku k\u00ebto veprime mund t\u00eb duken shum\u00eb t\u00eb ndryshme p\u00ebr nx\u00ebn\u00ebs t\u00eb ndrysh\u00ebm. Ai kaloi nj\u00eb m\u00ebsim: \u00e7far\u00eb do t\u00eb b\u00ebj\u00eb m\u00eb pas? Do t\u00eb filloj\u00eb detyr\u00ebn e sht\u00ebpis\u00eb? Do t\u00eb nis\u00eb aplikacionin celular? Do t\u00eb ndryshoj\u00eb kursin, do t\u00eb k\u00ebrkoj\u00eb t\u00eb ndryshoj\u00eb m\u00ebsuesin? Do t\u00eb hyj\u00eb menj\u00ebher\u00eb n\u00eb m\u00ebsimin tjet\u00ebr? Apo do t\u00eb largohet i zhg\u00ebnjyer? A \u00ebsht\u00eb e mundur t\u00eb identifikojm\u00eb, duke analizuar k\u00ebt\u00eb hart\u00eb, rregulla q\u00eb \u00e7ojn\u00eb n\u00eb p\u00ebrfundimin e suksessh\u00ebm t\u00eb kursit apo nga ana tjet\u00ebr, \u201elargimin\u201c e studentit?<\/p>\n

\"Si
\n
\nZakonisht p\u00ebr nd\u00ebrtimin e CJM p\u00ebrdoren mjete t\u00eb specializuara, shpeshher\u00eb shum\u00eb t\u00eb shtrenjta me kod t\u00eb mbyllur. Por ne doja t\u00eb shpikim di\u00e7ka t\u00eb thjesht\u00eb, q\u00eb t\u00eb k\u00ebrkonte p\u00ebrpjekje minimale dhe t\u00eb ishte sa m\u00eb shum\u00eb open-source. K\u00ebshtu nd\u00ebrtojm\u00eb iden\u00eb p\u00ebr t\u00eb p\u00ebrdorur zinxhir\u00ebt Markov \u2014 dhe ia dol\u00ebm. Krijuam nj\u00eb hart\u00eb, interpretuam t\u00eb dh\u00ebnat mbi sjelljen e student\u00ebve n\u00eb form\u00ebn e nj\u00eb grafiku, dhe pam\u00eb p\u00ebrgjigje krejt t\u00eb papritura p\u00ebr pyetje globale t\u00eb biznesit dhe madje gjet\u00ebm defekte t\u00eb thella t\u00eb fshehura. T\u00eb gjitha k\u00ebto i b\u00ebm\u00eb me ndihm\u00ebn e zgjidhjeve open-source t\u00eb skenar\u00ebve Python. N\u00eb k\u00ebt\u00eb artikull do t\u00eb flas p\u00ebr dy raste me k\u00ebto rezultate krejt t\u00eb papritura dhe do t\u00eb ndaj skenarin me t\u00eb gjith\u00eb ata q\u00eb jan\u00eb t\u00eb interesuar.
\n<\/a><\/noindex>
\nK\u00ebshtu, zinxhir\u00ebt Markov tregojn\u00eb probabilitetin e kalimeve midis ngjarjeve. Ja nj\u00eb shembull primitiv nga Wikipedia:<\/p>\n

\"Si
\n
\nK\u00ebtu \"E\" dhe \"A\" jan\u00eb ngjarje, shigjetat jan\u00eb kalimet midis tyre (p\u00ebrfshir\u00eb kalimin nga nj\u00eb ngjarje n\u00eb vetvete), dhe peshat e shigjetave jan\u00eb probabiliteti i kalimit (\"grafik i orientuar me pesha\").<\/p>\n

\u00c7far\u00eb p\u00ebrdor\u00ebm<\/h2>\n

\nZinxhiri u st\u00ebrvit me funksionalitetin standard t\u00eb Python, t\u00eb cilit iu dhan\u00eb loget e aktivitetit t\u00eb student\u00ebve. Grafiku n\u00eb matric\u00ebn e marr\u00eb u nd\u00ebrtua me bibliotek\u00ebn NetworkX.<\/p>\n

Logu duket k\u00ebshtu:<\/p>\n

\"Si
\n
\nKy \u00ebsht\u00eb nj\u00eb skedar csv q\u00eb p\u00ebrmban nj\u00eb tabel\u00eb me tre kolona: id e studentit, emri i ngjarjes, koha kur ndodhi ajo. K\u00ebto tri fusha jan\u00eb t\u00eb mjaftueshme p\u00ebr t\u00eb ndjekur l\u00ebvizjet e klientit, p\u00ebr t\u00eb nd\u00ebrtuar hart\u00ebn dhe n\u00eb fund p\u00ebr t\u00eb marr\u00eb zinxhirin Markov.<\/p>\n

Biblioteka kthen grafet e nd\u00ebrtuara n\u00eb formatin .dot ose .gexf. P\u00ebr vizualizimin e t\u00eb par\u00ebve mund t\u00eb p\u00ebrdorim paket\u00ebn falas Graphviz (instrumenti gvedit), ne kemi punuar me .gexf dhe Gephi, gjithashtu falas.<\/p>\n

Tani do t\u00eb sjell dy shembuj t\u00eb p\u00ebrdorimit t\u00eb zinxhir\u00ebve Markov, t\u00eb cilat na ndihmuan t\u00eb shohim ndryshe q\u00ebllimet tona, proceset e t\u00eb m\u00ebsuarit dhe vet\u00eb ekosistemin Skyeng. Po ashtu, rregulluam disa gabime.<\/p>\n

Shembulli i par\u00eb: aplikacioni mobil<\/h2>\n

\nFillimisht, ne hulumtuam rrug\u00ebn e studentit n\u00eb produktin ton\u00eb m\u00eb t\u00eb njohur \u2014 kursin General. N\u00eb at\u00eb koh\u00eb un\u00eb punoja n\u00eb departamentin e f\u00ebmij\u00ebve t\u00eb Skyeng dhe d\u00ebshironim t\u00eb shihnim sa efektivisht funksionon aplikacioni mobil me audienc\u00ebn ton\u00eb f\u00ebminore.<\/p>\n

Merrni logjet dhe kaloni ato p\u00ebrmes skriptit, kam marr\u00eb di\u00e7ka t\u00eb till\u00eb:<\/p>\n

\"Si<\/a><\/noindex><\/p>\n

Nod\u00ebn fillestare \u2014 Start General, dhe m\u00eb posht\u00eb tre nodet e daljes: studenti \u00abka fjetur\u00bb, ndryshoi kursin, p\u00ebrfundoi kursin.<\/p>\n