<\/a><\/noindex>Porosia e Data Science nuk \u00ebsht\u00eb thjesht t\u00eb kryesh detyra. Q\u00ebllimi \u00ebsht\u00eb t\u00eb studiohet dhe t\u00eb zhvillohen mund\u00ebsi t\u00eb reja t\u00eb forta p\u00ebr biznesin. Produktet dhe sh\u00ebrbimet algorithmike, si sistemet e rekomandimeve, nd\u00ebrveprimi me klient\u00ebt, klasifikimi i preferencave p\u00ebr stilin, p\u00ebrshtatja e madh\u00ebsive, dizajni i veshjeve, optimizimi i logjistik\u00ebs, zbulimi i trendeve sezonale dhe shum\u00eb t\u00eb tjera nuk mund t\u00eb zhvillohen paraprakisht. Ato duhet t\u00eb studiohen. Nuk ka asnj\u00eb skem\u00eb p\u00ebr riprodhim; k\u00ebto jan\u00eb mund\u00ebsi t\u00eb reja me pasigurin\u00eb q\u00eb i karakterizon. Koeficient\u00ebt, modelet, llojet e modeleve, hiperparametrat, t\u00eb gjitha elementet e nevojshme duhet t\u00eb studiohen p\u00ebrmes eksperimenteve, provave dhe gabimeve, si dhe p\u00ebrs\u00ebritjes. Me pin\u00eb, m\u00ebsimi dhe dizajni b\u00ebhen p\u00ebrpara, para momentit t\u00eb prodhimit. Me Data Science, ju m\u00ebsoni gjat\u00eb procesit, jo paraprakisht.<\/p>\n
N\u00eb fabrik\u00ebn e thikave, kur trajnimi \u00ebsht\u00eb n\u00eb vend t\u00eb par\u00eb, ne nuk presim dhe nuk duam q\u00eb punonj\u00ebsit t\u00eb improvizojn\u00eb mbi asnj\u00eb ve\u00e7ori t\u00eb produktit, p\u00ebrve\u00e7 p\u00ebr t\u00eb rritur efikasitetin e prodhimit. Specializimi i detyrave ka kuptim, pasi \u00e7on n\u00eb efikasitetin e proceseve dhe n\u00eb koherenc\u00ebn e prodhimit (pa b\u00ebr\u00eb ndryshime n\u00eb produktin p\u00ebrfundimtar).<\/p>\n
Por kur produkti ende \u00ebsht\u00eb n\u00eb zhvillim dhe q\u00ebllimi \u00ebsht\u00eb trajnimi, specializimi pengon q\u00ebllimet tona n\u00eb rastet e m\u00ebposhtme:<\/p>\n
1. Kjo rrit kostot e koordinimit. <\/b><\/p>\n
Pra k\u00ebt\u00eb, kemi shpenzimet e akumuluara gjat\u00eb koh\u00ebs s\u00eb kaluar n\u00eb komunikim, diskutim, arsyetim dhe p\u00ebrcaktimin e prioriteteve p\u00ebr pun\u00ebn q\u00eb duhet t\u00eb kryhet. K\u00ebto shpenzime rriten n\u00eb m\u00ebnyr\u00eb jo lineare me numrin e njer\u00ebzve t\u00eb p\u00ebrfshir\u00eb. (Si\u00e7 na m\u00ebsoi J. Richard Hackman, numri i marr\u00ebdh\u00ebnieve r rritet n\u00eb p\u00ebrputhje me funksionin e numrit t\u00eb an\u00ebtar\u00ebve n sipas k\u00ebsaj formule: r = (n ^ 2 - n) \/ 2. \u00c7do marr\u00ebdh\u00ebnie zbulon nj\u00eb sasi t\u00eb caktuar t\u00eb raportit t\u00eb shpenzimeve). Kur specialist\u00ebt e analiz\u00ebs s\u00eb t\u00eb dh\u00ebnave organizohen sipas funksioneve, n\u00eb \u00e7do hap, me \u00e7do ndryshim, me \u00e7do kalim sh\u00ebrbimi, etj., k\u00ebrkohet nj\u00eb num\u00ebr i madh specialist\u00ebsh, \u00e7ka rrit shpenzimet p\u00ebr koordinim. P\u00ebr shembull, specialist\u00ebt e modelimit statistik q\u00eb duan t\u00eb eksperimentojn\u00eb me funksione t\u00eb reja duhet t\u00eb koordinojn\u00eb veprimet e tyre me inxhinier\u00ebt e p\u00ebrpunimit t\u00eb t\u00eb dh\u00ebnave, t\u00eb cil\u00ebt p\u00ebrmir\u00ebsojn\u00eb sete t\u00eb dh\u00ebnash \u00e7do her\u00eb q\u00eb duan t\u00eb provojn\u00eb di\u00e7ka t\u00eb re. Po k\u00ebshtu, \u00e7do model i ri i trajnuar do t\u00eb thot\u00eb q\u00eb zhvilluesi i modelit do t\u00eb ket\u00eb nevoj\u00eb p\u00ebr dik\u00eb me t\u00eb cilin t\u00eb koordinoj\u00eb veprimet p\u00ebr ta vendosur at\u00eb n\u00eb veprim. Shpenzimet p\u00ebr koordinim sh\u00ebrbejn\u00eb si nj\u00eb tarif\u00eb p\u00ebr iterimin, duke i b\u00ebr\u00eb ato m\u00eb t\u00eb v\u00ebshtira dhe m\u00eb t\u00eb shtrenjta, dhe duke rritur mund\u00ebsin\u00eb p\u00ebr t'u t\u00ebrhequr nga k\u00ebrkimi. Kjo mund t\u00eb pengoj\u00eb m\u00ebsimin.<\/p>\n
2. Kjo e komplikon koh\u00ebn e pritjes. <\/b><\/p>\n
M\u00eb e frikshme se kostot e koordinimit \u00ebsht\u00eb koha e humbur midis ndryshimeve t\u00eb pun\u00ebs. Nd\u00ebrsa kostot e koordinimit zakonisht maten n\u00eb or\u00eb: koha q\u00eb nevojitet p\u00ebr t\u00eb mbajtur takime, diskutime, rishikime t\u00eb projekteve, koha e pritjes p\u00ebr zakonisht matet n\u00eb dit\u00eb, jav\u00eb ose madje muaj! Oraret e specialist\u00ebve jan\u00eb t\u00eb v\u00ebshtira p\u00ebr t'u p\u00ebrputhur, pasi \u00e7do specialist duhet t\u00eb jet\u00eb i shp\u00ebrndar\u00eb n\u00eb disa projekte. Nj\u00eb takim nj\u00eb or\u00ebsh p\u00ebr t\u00eb diskutuar ndryshimet mund t\u00eb zgjas\u00eb disa jav\u00eb p\u00ebr t\u00eb rregulluar fluksin e pun\u00ebs. Dhe pasi t\u00eb bien dakord p\u00ebr ndryshimet, duhet t\u00eb planifikohet edhe puna e v\u00ebrtet\u00eb n\u00eb kontekstin e shum\u00eb projekteve t\u00eb tjera q\u00eb z\u00ebn\u00eb koh\u00ebn e specialist\u00ebve. Puna q\u00eb ka t\u00eb b\u00ebj\u00eb me rregullimin e kodit ose hulumtimin, p\u00ebr t\u00eb cilat nevojiten vet\u00ebm disa or\u00eb ose dit\u00eb p\u00ebr t'u p\u00ebrfunduar, mund t\u00eb zgjas\u00eb shum\u00eb m\u00eb tep\u00ebr para se burimet t\u00eb b\u00ebhen t\u00eb disponueshme. Deri at\u00ebher\u00eb, iterimi dhe m\u00ebsimi ndalen.<\/p>\n
3. Kjo e ngushton kontekstin.<\/b><\/p>\n
Nd\u00ebrsa ndarja e pun\u00ebs mund t\u00eb kufizoj\u00eb m\u00ebsimin artificialisht, duke shp\u00ebrblyer njer\u00ebzit p\u00ebr q\u00ebndrimin n\u00eb specializimin e tyre. P\u00ebr shembull, nj\u00eb shkenc\u00ebtar hulumtues, q\u00eb duhet t\u00eb q\u00ebndroj\u00eb brenda funksionalitetit t\u00eb tij, do t\u00eb p\u00ebrqendroj\u00eb energjin\u00eb e tij n\u00eb eksperimente me lloje t\u00eb ndryshme algoritmesh: regresion, rrjeta neuronale, random forest dhe k\u00ebshtu me radh\u00eb. Sigurisht, nj\u00eb zgjedhje e mir\u00eb algoritmi mund t\u00eb sjell\u00eb p\u00ebrmir\u00ebsime t\u00eb ngadalta, por, zakonisht, shum\u00eb m\u00eb tep\u00ebr mund t\u00eb nxirret nga aktivitete t\u00eb tjera, t\u00eb tilla si integrimi i burimeve t\u00eb reja t\u00eb t\u00eb dh\u00ebnave. Po ashtu, kjo do t\u00eb ndihmoj\u00eb n\u00eb zhvillimin e nj\u00eb modeli q\u00eb p\u00ebrdor \u00e7do cop\u00eb shpjegueshm\u00ebrie t\u00eb pranishme n\u00eb t\u00eb dh\u00ebnat. Megjithat\u00eb, pika e saj e fort\u00eb mund t\u00eb jet\u00eb n\u00eb ndryshimin e funksionit t\u00eb synuar ose n\u00eb leht\u00ebsimin e disa kufizimeve. Kjo \u00ebsht\u00eb e v\u00ebshtir\u00eb p\u00ebr t'u par\u00eb ose b\u00ebr\u00eb, kur puna e saj \u00ebsht\u00eb e kufizuar. Pasi shkenc\u00ebtari specialist specializohet n\u00eb optimizimin e algoritmeve, ai ka shum\u00eb m\u00eb pak shanse t\u00eb angazhohet n\u00eb di\u00e7ka tjet\u00ebr, edhe n\u00ebse kjo sjell dobi t\u00eb konsiderueshme.<\/p>\n
Le t\u00eb shpjegojm\u00eb shenjat q\u00eb shfaqen kur ekipet e data science funksionojn\u00eb si fabrika kunjsh (p\u00ebr shembull, n\u00eb p\u00ebrdit\u00ebsime t\u00eb thjeshta statusi): \"pritja e ndryshimeve n\u00eb kanalin e t\u00eb dh\u00ebnave\" dhe \"pritja e burimeve t\u00eb ML Eng\", t\u00eb cilat jan\u00eb bllokues t\u00eb zakonsh\u00ebm. Megjithat\u00eb, un\u00eb besoj se ndikimi m\u00eb i rreziksh\u00ebm \u00ebsht\u00eb ai \u00e7far\u00eb nuk v\u00ebreni, sepse nuk mund t\u00eb pendoheni p\u00ebr at\u00eb q\u00eb nuk e dini. Nj\u00eb p\u00ebrmbushje e p\u00ebrsosur e k\u00ebrkesave dhe qet\u00ebsimi i vet\u00eb-p\u00ebrmbushur, arritur si pasoj\u00eb e efikasitetit t\u00eb proceseve, mund t\u00eb fsheh\u00eb t\u00eb v\u00ebrtet\u00ebn se organizatat nuk jan\u00eb t\u00eb njohura me p\u00ebrfitimet e m\u00ebsimit q\u00eb po i humbin.<\/p>\n
Zgjidhja e k\u00ebtij problemi, padyshim, \u00ebsht\u00eb heqja e metod\u00ebs s\u00eb fabrik\u00ebs s\u00eb thikave. P\u00ebr t\u00eb stimuluar m\u00ebsimin dhe iteracionin, rolet n\u00eb shkenc\u00ebn e t\u00eb dh\u00ebnave duhet t\u00eb jen\u00eb t\u00eb p\u00ebrgjithshme, por me detyra t\u00eb gjera q\u00eb nuk varen nga funksioni teknik, dometh\u00ebn\u00eb organizoni specialist\u00ebt e t\u00eb dh\u00ebnave n\u00eb nj\u00eb m\u00ebnyr\u00eb q\u00eb t\u00eb optimizohen p\u00ebr m\u00ebsim. Kjo do t\u00eb thot\u00eb se \u00ebsht\u00eb e nevojshme t\u00eb pun\u00ebsohen \"specialist\u00eb me diapazon t\u00eb plot\u00eb\" \u2014 specialist\u00eb t\u00eb p\u00ebrgjithsh\u00ebm q\u00eb mund t\u00eb kryejn\u00eb funksione t\u00eb ndryshme: nga konceptimi deri te modelimi, nga implementimi deri te matja. \u00cbsht\u00eb e r\u00ebnd\u00ebsishme t\u00eb theksohet se nuk supozoj se gjat\u00eb pun\u00ebsimit t\u00eb specialist\u00ebve me diapazon t\u00eb plot\u00eb, numri i punonj\u00ebsve duhet t\u00eb zvogelohet. M\u00eb shum\u00eb do t\u00eb supozoj se kur ata organizohen ndryshe, stimujt e tyre lidhen m\u00eb mir\u00eb me p\u00ebrfitimet nga m\u00ebsimi dhe efektiviteti. P\u00ebr shembull, ju keni nj\u00eb ekip prej tre personash q\u00eb ka tri cil\u00ebsi t\u00eb biznesit. N\u00eb fabrik\u00ebn e prodhimit t\u00eb thikave, \u00e7do specialist do t'i kushtonte nj\u00eb t\u00eb tret\u00ebn e koh\u00ebs \u00e7do detyre profesionale, pasi askush tjet\u00ebr nuk do t\u00eb ishte n\u00eb gjendje t\u00eb kryej\u00eb pun\u00ebn e tij. N\u00eb nj\u00eb diapazon t\u00eb plot\u00eb, \u00e7do punonj\u00ebs universal \u00ebsht\u00eb plot\u00ebsisht i angazhuar n\u00eb t\u00eb gjith\u00eb procesin e biznesit, rritjen e shkall\u00ebs s\u00eb pun\u00ebs dhe m\u00ebsimin.<\/p>\n
Me fewer people supporting the production cycle, coordination decreases. The universal smoothly transitions between functions, expanding the data pipeline to add more data, experimenting with new features in models, deploying new versions in production for causal measurements, and repeating steps as quickly as new ideas emerge. Of course, the universal performs different functions sequentially, not in parallel. After all, it's just one person. However, completing a task typically takes only a small portion of the time required to access another specialized resource. Thus, iteration time decreases.<\/p>\n
Universali yn\u00eb ndoshta nuk \u00ebsht\u00eb aq i aft\u00eb sa nj\u00eb specialist n\u00eb nj\u00eb funksion t\u00eb caktuar pune, por ne nuk k\u00ebrkojm\u00eb p\u00ebrsosm\u00ebri funksionale ose p\u00ebrmir\u00ebsime t\u00eb vogla t\u00eb vazhdueshme. P\u00ebrkundrazi, ne k\u00ebrkojm\u00eb t\u00eb studiojm\u00eb dhe t\u00eb zgjerojm\u00eb vazhdimisht detyrat profesionale me ndikim gradual. Me nj\u00eb kontekst t\u00eb plot\u00eb p\u00ebr nj\u00eb zgjidhje t\u00eb plot\u00eb, ai sheh mund\u00ebsi q\u00eb nj\u00eb specialist i ngusht\u00eb do t'i injoronte. Ai ka m\u00eb shum\u00eb ide dhe m\u00eb shum\u00eb mund\u00ebsi. Ai gjithashtu d\u00ebshton. Megjithat\u00eb, \u00e7mimi i d\u00ebshtimit \u00ebsht\u00eb i ul\u00ebt, nd\u00ebrsa p\u00ebrfitimet e m\u00ebsimit jan\u00eb t\u00eb m\u00ebdha. Kjo asimetri mb\u00ebshtet iterimin e shpejt\u00eb dhe shp\u00ebrblen m\u00ebsimin.<\/p>\n
\u00cbsht\u00eb e r\u00ebnd\u00ebsishme t\u00eb theksohet se shkalla e autonomis\u00eb dhe larmis\u00eb s\u00eb aft\u00ebsive t\u00eb ofruara nga shkenc\u00ebtar\u00ebt q\u00eb punojn\u00eb me grupe t\u00eb plota, n\u00eb mas\u00eb t\u00eb madhe varet nga besueshm\u00ebria e platform\u00ebs s\u00eb t\u00eb dh\u00ebnave mbi t\u00eb cil\u00ebn ata mund t\u00eb punojn\u00eb. Nj\u00eb platform\u00eb t\u00eb dh\u00ebnash e nd\u00ebrtuar mir\u00eb e heq pesh\u00ebn nga shkenc\u00ebtar\u00ebt q\u00eb merren me p\u00ebrpunimin e t\u00eb dh\u00ebnave, duke i mbrojtur ata nga kompleksiteti i konteinerizimit, p\u00ebrpunimit t\u00eb shp\u00ebrndar\u00eb, kalimeve automatike n\u00eb burime t\u00eb tjera dhe koncepteve t\u00eb tjera p\u00ebrparimtare kompjuterike. P\u00ebrve\u00e7 abstrahimit, nj\u00eb platform\u00eb t\u00eb dh\u00ebnash e besueshme mund t\u00eb siguroj\u00eb lidhje pa pengesa me infrastruktur\u00ebn eksperimentale, t\u00eb automatizoj\u00eb monitorimin dhe sistemin e njoftimit, t\u00eb ofroj\u00eb auto-skallinim dhe vizualizim t\u00eb rezultateve algoritmike dhe t\u00eb depurimit. K\u00ebto komponent\u00eb jan\u00eb projektuar dhe nd\u00ebrtuar nga inxhinier\u00ebt e platform\u00ebs s\u00eb t\u00eb dh\u00ebnave, dometh\u00ebn\u00eb nuk kalohen nga specialisti n\u00eb shkenc\u00ebn e t\u00eb dh\u00ebnave p\u00ebr ekipin e zhvilluesve t\u00eb platform\u00ebs s\u00eb t\u00eb dh\u00ebnave. \u00cbsht\u00eb specialisti n\u00eb shkenc\u00ebn e t\u00eb dh\u00ebnave q\u00eb \u00ebsht\u00eb p\u00ebrgjegj\u00ebs p\u00ebr t\u00eb gjith\u00eb kodin q\u00eb p\u00ebrdoret p\u00ebr t\u00eb drejtuar platform\u00ebn.<\/p>\n
Edhe mua nj\u00ebher\u00eb m\u00eb \u00ebsht\u00eb interesuar ndarja funksionale e pun\u00ebs duke p\u00ebrdorur efikasitetin e proceseve, por me metod\u00ebn e provave dhe gabimeve (nuk ka m\u00ebnyr\u00eb m\u00eb t\u00eb mir\u00eb p\u00ebr t\u00eb m\u00ebsuar) kam zbuluar se rollet tipike kontribuojn\u00eb m\u00eb mir\u00eb n\u00eb m\u00ebsim dhe inovacion dhe ofrojn\u00eb tregues t\u00eb sakt\u00eb: identifikimin dhe nd\u00ebrtimin e nj\u00eb numri shum\u00eb m\u00eb t\u00eb madh mund\u00ebsish biznesi se sa qasja e specializuar. (Nj\u00eb m\u00ebnyr\u00eb m\u00eb efikase p\u00ebr t\u00eb m\u00ebsuar p\u00ebr k\u00ebt\u00eb qasje n\u00eb organizim, p\u00ebrve\u00e7 metod\u00ebs s\u00eb provave dhe gabimeve p\u00ebrmes s\u00eb cil\u00ebs kam kaluar, \u00ebsht\u00eb t\u00eb lexosh librin e Amy Edmondson 'Bashk\u00ebpunimi n\u00eb ekip: si m\u00ebsojn\u00eb organizatat, krijojn\u00eb inovacione dhe konkurrojn\u00eb n\u00eb ekonomin\u00eb e njohurive').<\/p>\n
Ka disa supozita t\u00eb r\u00ebnd\u00ebsishme q\u00eb mund ta b\u00ebjn\u00eb k\u00ebt\u00eb qasje n\u00eb organizim m\u00eb t\u00eb besueshme ose m\u00eb pak t\u00eb besueshme n\u00eb disa kompani. Procesi i iteracionit ul kostot e provave dhe gabimeve. N\u00ebse \u00e7mimi i gabimit \u00ebsht\u00eb i lart\u00eb, mund t\u00eb d\u00ebshironit t'i reduktoni ata (por, kjo nuk rekomandohet p\u00ebr aplikimet mjek\u00ebsore ose prodhimin). P\u00ebr m\u00eb tep\u00ebr, n\u00ebse po merrni me petabajt ose eksabajt t\u00eb dh\u00ebnash, mund t\u00eb nevojitet specializim n\u00eb projektimin e t\u00eb dh\u00ebnave. Po ashtu, n\u00ebse mbajtja e mund\u00ebsive t\u00eb biznesit n\u00eb rrjet dhe disponueshm\u00ebria e tyre \u00ebsht\u00eb m\u00eb e r\u00ebnd\u00ebsishme se p\u00ebrmir\u00ebsimi i tyre, ekselenca funksionale mund t\u00eb tejkaloj\u00eb m\u00ebsimin. S\u00eb fundi, modeli i stack-ut t\u00eb plot\u00eb mb\u00ebshtetet n\u00eb mendimet e njer\u00ebzve q\u00eb din\u00eb p\u00ebr k\u00ebt\u00eb. Ata nuk jan\u00eb nj\u00eblloj; mund t\u00eb gjenden ose t\u00eb trajnohen vet\u00eb. Megjithat\u00eb, ata jan\u00eb t\u00eb k\u00ebrkuar shum\u00eb, dhe p\u00ebr t\u00eb t\u00ebrhequr dhe mbajtur ata n\u00eb kompani do t\u00eb k\u00ebrkohet nj\u00eb kompensim material konkurrues, vlera t\u00eb q\u00ebndrueshme korporative dhe pun\u00eb interesante. Sigurohuni q\u00eb kultura juaj korporative mund t\u00eb ofroj\u00eb kushte t\u00eb tilla.<\/p>\n
Edhe me gjith\u00e7ka t\u00eb th\u00ebn\u00eb, mendoj se modeli i plot\u00eb i stack-ut ofron kushtet m\u00eb t\u00eb mira p\u00ebr fillim. Filloni me ta, dhe pastaj l\u00ebvizni me vet\u00ebdije drejt ndarjes funksionale t\u00eb pun\u00ebs vet\u00ebm kur kjo \u00ebsht\u00eb me t\u00eb v\u00ebrtet\u00eb e nevojshme.<\/p>\n
Ekzistojn\u00eb edhe disavantazhe t\u00eb tjera t\u00eb specializimit funksional. Kjo mund t\u00eb \u00e7oj\u00eb n\u00eb humbje t\u00eb p\u00ebrgjegj\u00ebsis\u00eb dhe pasivitetit nga ana e punonj\u00ebsve. Edhe Smith e kritikon ndarjen e pun\u00ebs, duke sugjeruar se ajo \u00e7on n\u00eb prishjen e talentit, q\u00eb do t\u00eb thot\u00eb se punonj\u00ebsit b\u00ebhen t\u00eb paditur dhe t\u00eb mbyllur, pasi rolet e tyre kufizohen n\u00eb disa detyra t\u00eb p\u00ebrs\u00ebritura. Nd\u00ebrsa specializimi mund t\u00eb siguroj\u00eb efikasitetin e procesit, ai rrall\u00ebher\u00eb frym\u00ebzon punonj\u00ebsit.<\/p>\n
Nga ana e role universale sigurojn\u00eb t\u00eb gjitha ato q\u00eb stimulojn\u00eb k\u00ebnaq\u00ebsin\u00eb n\u00eb pun\u00eb: autonomi, mjesht\u00ebri dhe q\u00ebllim. Autonomia lidhet me at\u00eb q\u00eb ata nuk varen nga di\u00e7ka n\u00eb arritjen e suksesit. Mjesht\u00ebria konsiston n\u00eb avantazhe t\u00eb forta konkurruese. Nd\u00ebrsa q\u00ebllimi p\u00ebrfshin mund\u00ebsin\u00eb e ndikimit n\u00eb biznesin q\u00eb ata krijojn\u00eb. N\u00ebse arrijm\u00eb t'i b\u00ebjm\u00eb njer\u00ebzit t\u00eb angazhohen n\u00eb pun\u00ebn e tyre dhe t\u00eb ken\u00eb nj\u00eb ndikim t\u00eb madh n\u00eb kompanin\u00eb, at\u00ebher\u00eb gjith\u00e7ka tjet\u00ebr do t\u00eb vendoset n\u00eb vendin e vet.habr.com<\/a><\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"HIROSHI WATANABE\/GETTY IMAGES \u0412 \u043a\u043d\u0438\u0433\u0435 \u00ab\u0411\u043e\u0433\u0430\u0442\u0441\u0442\u0432\u043e \u043d\u0430\u0440\u043e\u0434\u043e\u0432\u00bb \u0410\u0434\u0430\u043c \u0421\u043c\u0438\u0442 \u043f\u043e\u043a\u0430\u0437\u044b\u0432\u0430\u0435\u0442, \u043a\u0430\u043a \u0440\u0430\u0437\u0434\u0435\u043b\u0435\u043d\u0438\u0435 \u0442\u0440\u0443\u0434\u0430 \u0441\u0442\u0430\u043d\u043e\u0432\u0438\u0442\u0441\u044f \u0433\u043b\u0430\u0432\u043d\u044b\u043c \u0438\u0441\u0442\u043e\u0447\u043d\u0438\u043a\u043e\u043c \u043f\u043e\u0432\u044b\u0448\u0435\u043d\u0438\u044f \u043f\u0440\u043e\u0438\u0437\u0432\u043e\u0434\u0438\u0442\u0435\u043b\u044c\u043d\u043e\u0441\u0442\u0438. \u041f\u0440\u0438\u043c\u0435\u0440\u043e\u043c \u0432\u044b\u0441\u0442\u0443\u043f\u0430\u0435\u0442 \u043b\u0438\u043d\u0438\u044f \u0441\u0431\u043e\u0440\u043a\u0438 \u0437\u0430\u0432\u043e\u0434\u0430 \u043f\u043e \u043f\u0440\u043e\u0438\u0437\u0432\u043e\u0434\u0441\u0442\u0432\u0443 \u0431\u0443\u043b\u0430\u0432\u043e\u043a: \u00ab\u041e\u0434\u0438\u043d \u0440\u0430\u0431\u043e\u0447\u0438\u0439 \u0442\u044f\u043d\u0435\u0442 \u043f\u0440\u043e\u0432\u043e\u043b\u043e\u043a\u0443, \u0434\u0440\u0443\u0433\u043e\u0439 \u0432\u044b\u043f\u0440\u044f\u043c\u043b\u044f\u0435\u0442 \u0435\u0451, \u0442\u0440\u0435\u0442\u0438\u0439 \u043e\u0431\u0440\u0435\u0437\u0430\u0435\u0442, \u0447\u0435\u0442\u0432\u0435\u0440\u0442\u044b\u0439 \u0437\u0430\u043e\u0441\u0442\u0440\u044f\u0435\u0442 \u043a\u043e\u043d\u0435\u0446, \u043f\u044f\u0442\u044b\u0439 \u043e\u0431\u0442\u0430\u0447\u0438\u0432\u0430\u0435\u0442 \u0434\u0440\u0443\u0433\u043e\u0439 \u043a\u043e\u043d\u0435\u0446 \u0434\u043b\u044f \u043d\u0430\u0441\u0430\u0436\u0438\u0432\u0430\u043d\u0438\u044f \u0433\u043e\u043b\u043e\u0432\u043a\u0438\u00bb. \u0411\u043b\u0430\u0433\u043e\u0434\u0430\u0440\u044f \u0441\u043f\u0435\u0446\u0438\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u0438, \u043e\u0440\u0438\u0435\u043d\u0442\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u043e\u0439 \u043d\u0430 \u043e\u043f\u0440\u0435\u0434\u0435\u043b\u0435\u043d\u043d\u044b\u0435 \u0444\u0443\u043d\u043a\u0446\u0438\u0438, \u043a\u0430\u0436\u0434\u044b\u0439 \u0440\u0430\u0431\u043e\u0442\u043d\u0438\u043a \u0441\u0442\u0430\u043d\u043e\u0432\u0438\u0442\u0441\u044f \u0432\u044b\u0441\u043e\u043a\u043e\u043a\u0432\u0430\u043b\u0438\u0444\u0438\u0446\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u044b\u043c […]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":1,"featured_media":24671,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[702],"tags":[],"class_list":["post-32907","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-novosti-interneta"],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t
![\"Pse](\"\/wp-content\/uploads\/2019\/05\/19d1e32e5d57abd463376af240cfa42e.jpg\")