{"id":12283,"date":"2023-05-28T14:00:52","date_gmt":"2023-05-28T11:00:52","guid":{"rendered":"https:\/\/bestessayhomework.com\/tr\/?p=12283"},"modified":"2023-05-28T14:00:52","modified_gmt":"2023-05-28T11:00:52","slug":"guc-yasasi-dagilimlari","status":"publish","type":"post","link":"https:\/\/bestessayhomework.com\/tr\/guc-yasasi-dagilimlari\/","title":{"rendered":"G\u00fc\u00e7 Yasas\u0131 Da\u011f\u0131l\u0131mlar\u0131"},"content":{"rendered":"

Power-Law Al\u0131nt\u0131 Da\u011f\u0131t\u0131mlar\u0131\u00a0<\/span><\/strong><\/h3>\n

Bilimsel makalelere yap\u0131lan at\u0131flar\u0131n istatistiksel da\u011f\u0131l\u0131mlar\u0131n\u0131n zaman i\u00e7indeki geli\u015fimini analiz ediyoruz. Bu da\u011f\u0131l\u0131mlar kuvvet yasas\u0131 ba\u011f\u0131ml\u0131l\u0131\u011f\u0131n\u0131 takip ediyor gibi g\u00f6r\u00fcnse de dura\u011fan olmad\u0131klar\u0131n\u0131 ve kuvvet yasas\u0131 uyumunun \u00fcss\u00fcn\u00fcn zamanla azald\u0131\u011f\u0131n\u0131 ve doyuma ula\u015fmad\u0131\u011f\u0131n\u0131 g\u00f6r\u00fcyoruz.<\/span><\/p>\n

At\u0131f da\u011f\u0131l\u0131mlar\u0131n\u0131n dura\u011fan olmamas\u0131n\u0131, d\u00fc\u015f\u00fck at\u0131fl\u0131 ve y\u00fcksek at\u0131fl\u0131 makalelerin farkl\u0131 uzun \u00f6m\u00fcrl\u00fcl\u00fcklerine ba\u011fl\u0131yoruz. Makalelerin at\u0131f y\u00f6r\u00fcngelerini \u00f6l\u00e7erek, d\u00fc\u015f\u00fck at\u0131fl\u0131 makalelerin at\u0131f kariyerlerinin 10-15 y\u0131l sonra doygunlu\u011fa ula\u015ft\u0131\u011f\u0131n\u0131, y\u00fcksek at\u0131fl\u0131 makalelerin ise s\u00fcresiz olarak artmaya devam etti\u011fini bulduk.<\/span><\/p>\n

Bir makalenin at\u0131f say\u0131s\u0131 belli bir at\u0131f e\u015fi\u011fini a\u015ft\u0131\u011f\u0131nda, makale bir ka\u00e7ak haline gelir. B\u00f6ylece, at\u0131f da\u011f\u0131l\u0131m\u0131n\u0131n bir g\u00fc\u00e7 yasas\u0131 ba\u011f\u0131ml\u0131l\u0131\u011f\u0131 gibi g\u00f6r\u00fcnebilmesine ra\u011fmen, \u00f6l\u00e7ekten ba\u011f\u0131ms\u0131z olmad\u0131\u011f\u0131n\u0131 ve ka\u00e7aklar\u0131n ba\u015flang\u0131c\u0131yla ili\u015fkili gizli bir dinamik \u00f6l\u00e7ek oldu\u011funu g\u00f6steriyoruz. Kopyalama\/y\u00f6nlendirme\/\u00fc\u00e7l\u00fc kapatmaya dayal\u0131 al\u0131nt\u0131 dinamikleri modelimizin bu sorunlar\u0131 olduk\u00e7a iyi a\u00e7\u0131klad\u0131\u011f\u0131n\u0131 g\u00f6steriyoruz.<\/span><\/p>\n

G\u00fc\u00e7 Yasas\u0131 Da\u011f\u0131l\u0131mlar\u0131<\/span><\/strong><\/h3>\n

Son derece \u00e7arp\u0131k istatistiksel da\u011f\u0131l\u0131mlar, bir as\u0131rdan daha uzun bir s\u00fcre \u00f6nce ke\u015ffedildi ve yo\u011fun ara\u015ft\u0131rma konusu olmaya devam ediyor. Bu kal\u0131n kuyruklu s\u00fcrekli da\u011f\u0131l\u0131mlardan en \u00f6nemlisi g\u00fc\u00e7 yasas\u0131d\u0131r. p(x)’in olas\u0131l\u0131k yo\u011funluk fonksiyonu oldu\u011fu yerde, \u03bd \u00fcs ve w kaymad\u0131r.<\/span><\/p>\n

Sonlu momentleri olan Gauss da\u011f\u0131l\u0131m\u0131n\u0131n aksine, kuvvet yasas\u0131 da\u011f\u0131l\u0131mlar\u0131n\u0131n momentleri birbirinden uzakla\u015fabilir. \u00d6zellikle \u03bd \u2264 2 i\u00e7in ortalama, \u03bd \u2264 3 i\u00e7in varyans \u0131raksar. B\u00f6ylece 2 \u2264 \u03bd \u2264 3 olan kuvvet yasas\u0131 da\u011f\u0131l\u0131m\u0131n\u0131n birinci ve ikinci momentleri kuyru\u011fu taraf\u0131ndan belirlenir ve bu nedenledir bu da\u011f\u0131l\u0131m, \u015fi\u015fman veya a\u011f\u0131r kuyruklu olarak adland\u0131r\u0131l\u0131r.<\/span><\/p>\n

Kuvvet yasas\u0131 ve log-normal da\u011f\u0131l\u0131mlar\u0131n kendine \u00f6zg\u00fc bir \u00f6zelli\u011fi \u00f6l\u00e7eksiz olmalar\u0131d\u0131r, yani olas\u0131l\u0131k yo\u011funluk fonksiyonlar\u0131 p(x) ve p(x\/S), burada S bir sabittir, ayn\u0131 \u015fekle sahiptir ve log-log \u00f6l\u00e7e\u011finde sadece kayd\u0131r\u0131l\u0131r, ba\u015fka bir deyi\u015fle, bu da\u011f\u0131l\u0131mlar veya en az\u0131ndan kuyruklar\u0131 kendine benzerdir.<\/span><\/p>\n

Olduk\u00e7a \u00e7arp\u0131k da\u011f\u0131l\u0131mlar\u0131n \u00f6l\u00e7eksiz \u00f6zelli\u011fi, fraktallar, faz ge\u00e7i\u015fleri ve kritik olgular gibi di\u011fer \u00f6l\u00e7eksiz olgularla derin analojiler arayan bir\u00e7ok fizik\u00e7i i\u00e7in bir hayranl\u0131k kayna\u011f\u0131 olmu\u015ftur.<\/span><\/p>\n

Da\u011f\u0131l\u0131mlar\u0131n Deneysel De\u011ferlendirmesi<\/span><\/strong><\/h3>\n

Karma\u015f\u0131k a\u011flardaki derece da\u011f\u0131l\u0131mlar\u0131n\u0131 hesaba katmak i\u00e7in, \u00f6zellikle at\u0131f da\u011f\u0131l\u0131mlar\u0131n\u0131n karakterizasyonu i\u00e7in, \u015fi\u015fman kuyruklu da\u011f\u0131l\u0131mlara s\u0131kl\u0131kla ba\u015fvurulur. Bunu yapman\u0131n en basit yolu, \u00f6l\u00e7\u00fclen derece da\u011f\u0131l\u0131m\u0131n\u0131 log-log \u00f6l\u00e7e\u011finde \u00e7izmektir.<\/span><\/p>\n

D\u00fcz \u00e7izgi kuvvet yasas\u0131n\u0131 g\u00f6sterirken, parabol log-normal da\u011f\u0131l\u0131m\u0131 \u00f6nerir. Pratikte, log-log \u00f6l\u00e7e\u011findeki e\u011frilik testi, bu iki da\u011f\u0131l\u0131m aras\u0131nda iyi bir ayr\u0131m yapmaz. Asl\u0131nda, log-normal da\u011f\u0131l\u0131m \u00e7ok geni\u015fse, o zaman b\u00fcy\u00fck bir parabol par\u00e7as\u0131 d\u00fcz bir \u00e7izgi gibi g\u00f6r\u00fcn\u00fcr.<\/span><\/p>\n

Kesikli da\u011f\u0131l\u0131mlar i\u00e7in durum daha da k\u00f6t\u00fcd\u00fcr, \u00e7\u00fcnk\u00fc ayr\u0131k kuvvet yasas\u0131 da\u011f\u0131l\u0131m\u0131n\u0131n log-log grafi\u011fi k\u00fc\u00e7\u00fck derecelerde d\u0131\u015fb\u00fckey \u015fekle sahiptir ve bu, bu da\u011f\u0131l\u0131m\u0131 lognormalden ay\u0131rt etme problemini daha da k\u00f6t\u00fcle\u015ftirir.<\/span><\/p>\n

\u0130statistiksel bir da\u011f\u0131l\u0131m\u0131n veya bir k\u0131sm\u0131n\u0131n log-log grafi\u011fi d\u00fcz bir \u00e7izgi gibi g\u00f6r\u00fcnse bile, e\u011fimini bulmak kolay bir i\u015f de\u011fildir. Asl\u0131nda, \u00e7o\u011fu da\u011f\u0131l\u0131m k\u00fc\u00e7\u00fck derecelerde yuvarland\u0131\u011f\u0131ndan, e\u011fimi \u00f6l\u00e7mek i\u00e7in k\u00fc\u00e7\u00fck dereceler kesilip da\u011f\u0131l\u0131m\u0131n kuyru\u011funa odaklan\u0131lmal\u0131d\u0131r.<\/span><\/p>\n

Bu kesme prosed\u00fcr\u00fc s\u00fcbjektiftir ve bir belirsizlik kayna\u011f\u0131d\u0131r. Kuvvet kanunu derece da\u011f\u0131l\u0131m\u0131n\u0131n deneysel olarak tan\u0131mlanmas\u0131n\u0131n zorluklar\u0131, karma\u015f\u0131k a\u011flardaki derece da\u011f\u0131l\u0131mlar\u0131n\u0131n kuvvet kanunu, log-normal veya uzat\u0131lm\u0131\u015f \u00fcstel taraf\u0131ndan daha iyi tan\u0131mlan\u0131p tan\u0131mlanmad\u0131\u011f\u0131 konusunda \u00f6nemli bir tart\u0131\u015fma yaratt\u0131.<\/span><\/p>\n

\n

G\u00fc\u00e7 yasas\u0131<\/a> da\u011f\u0131l\u0131m\u0131<\/span>
\npareto analizi 80\/20 kural\u0131 nedir<\/span>
\n80\/20 kural\u0131 \u00f6rnek<\/span>
\nPareto ilkesi \u00f6rnek<\/span>
\nPareto Prensibi zaman y\u00f6netimi<\/span>
\nPareto Kural\u0131 Nedir<\/span>
\nG\u00fc\u00e7 ve otorite aras\u0131ndaki ili\u015fki<\/span>
\nPareto etkisi Nedir<\/span><\/p>\n

\n

At\u0131f da\u011f\u0131l\u0131mlar\u0131n\u0131n de\u011ferlendirilmesinin tarihi, bu t\u00fcr tart\u0131\u015fmalara iyi bir \u00f6rnektir. \u00c7al\u0131\u015fmalardan ba\u015flayarak al\u0131nt\u0131lar, \u00fcs = 2.5\u20133.16 ile ayr\u0131 bir kuvvet yasas\u0131 da\u011f\u0131l\u0131m\u0131 ile yerle\u015ftirildi.<\/span><\/p>\n

Daha sonra moda, \u03c3 = 1\u20131.2 ile log-normal veya ayr\u0131kla\u015ft\u0131r\u0131lm\u0131\u015f log-normal uydurma i\u015flevine do\u011fru kayd\u0131. Son \u00e7al\u0131\u015fmalar, 3 ile 4 aras\u0131nda de\u011fi\u015fen \u00fcs \u03bd ile g\u00fc\u00e7 yasas\u0131 da\u011f\u0131l\u0131m\u0131n\u0131 tekrar iddia etti.<\/span><\/p>\n

Karma\u015f\u0131k bir a\u011f\u0131n i\u015flevsel derece da\u011f\u0131l\u0131m\u0131 bi\u00e7imini bulmak neden bu kadar \u00f6nemlidir? Bunun i\u00e7in motivasyon, bu fonksiyonel formun a\u011f b\u00fcy\u00fcme mekanizmas\u0131na dair bir ipucu oldu\u011fu inanc\u0131na dayan\u0131yordu. Bununla birlikte, derece da\u011f\u0131l\u0131m\u0131ndan b\u00fcy\u00fcme mekanizmas\u0131n\u0131 kesin olarak belirlemenin etkisiz ve belirsiz oldu\u011fu kan\u0131tlanm\u0131\u015ft\u0131r.<\/span><\/p>\n

Bu nedenle, a\u011f olu\u015fturma mekanizmas\u0131n\u0131 derece da\u011f\u0131l\u0131mlar\u0131ndan t\u00fcretme giri\u015fimlerini b\u0131rakmak ve bu mekanizmay\u0131 do\u011frudan, \u00f6rne\u011fin zaman serisi analizi uygulayarak \u00f6l\u00e7mek \u00f6nerilir.<\/span><\/p>\n

Biz de bu \u00f6neriyi yak\u0131ndan takip ettik ve at\u0131f a\u011f\u0131n\u0131n b\u00fcy\u00fcme mekanizmas\u0131n\u0131 mikroskobik \u00f6l\u00e7\u00fcmlerden belirledik. Bu mekanizma, bu \u00e7al\u0131\u015fman\u0131n \u00f6nceki b\u00f6l\u00fcmlerinde a\u00e7\u0131klanm\u0131\u015ft\u0131r. Bu b\u00f6l\u00fcmde \u015fu soruyu ele al\u0131yoruz: al\u0131nt\u0131 a\u011flar\u0131nda derece da\u011f\u0131l\u0131m\u0131n\u0131n i\u015flevsel bi\u00e7imi nedir? Bu soruya iki y\u00f6nden yakla\u015ft\u0131k.<\/span><\/p>\n

\u0130lk olarak, birka\u00e7 iyi tan\u0131mlanm\u0131\u015f at\u0131f a\u011f\u0131 se\u00e7tik. Her biri i\u00e7in derece da\u011f\u0131l\u0131mlar\u0131n\u0131n zamanla nas\u0131l geli\u015fti\u011fini \u00f6l\u00e7t\u00fck. Bu at\u0131f da\u011f\u0131l\u0131mlar\u0131n\u0131 yayg\u0131n olarak kabul edilen stratejiler kullanarak ampirik olarak analiz ettik ve dura\u011fan olmad\u0131klar\u0131n\u0131 ve uzun zaman s\u0131n\u0131r\u0131nda bile baz\u0131 s\u0131n\u0131rlay\u0131c\u0131 da\u011f\u0131l\u0131mlara yak\u0131nsamad\u0131klar\u0131n\u0131 bulduk.<\/span><\/p>\n

Bu dura\u011fan olmama, al\u0131nt\u0131 a\u011flar\u0131nda derece da\u011f\u0131l\u0131m\u0131n\u0131 karakterize etmeye y\u00f6nelik \u00f6nceki \u00e7abalar\u0131n neden bu kadar belirsiz oldu\u011funu a\u00e7\u0131kl\u0131yor. \u0130kinci olarak, bu al\u0131nt\u0131 da\u011f\u0131l\u0131mlar\u0131n\u0131, stokastik al\u0131nt\u0131 dinamikleri modelimizi kullanarak modelledik ve dura\u011fan olmama i\u00e7in a\u00e7\u0131klama bulduk.<\/span><\/p>\n

At\u0131f da\u011f\u0131l\u0131mlar\u0131n\u0131n dura\u011fan olmamas\u0131 ve bunlar\u0131n \u201cg\u00fc\u00e7 yasas\u0131\u201d \u015feklinin her ikisinin de do\u011frusal olmayan at\u0131f dinamiklerinden kaynakland\u0131\u011f\u0131 ortaya \u00e7\u0131kt\u0131. \u00d6zellikle, do\u011frusal olmama, al\u0131nt\u0131 a\u011flar\u0131n\u0131n art\u0131k \u00f6l\u00e7ekten ba\u011f\u0131ms\u0131z olarak kabul edilemeyece\u011fi \u015fekilde belirli bir \u00f6l\u00e7ek getirir.<\/span><\/p>\n

At\u0131f da\u011f\u0131l\u0131mlar\u0131n\u0131n i\u015flevsel bi\u00e7imini bulmak i\u00e7in, birka\u00e7 iyi tan\u0131mlanm\u0131\u015f ara\u015ft\u0131rma alan\u0131 se\u00e7tik ve ayn\u0131 y\u0131l yay\u0131nlanan t\u00fcm orijinal ara\u015ft\u0131rma makalelerine (genel bak\u0131\u015flar hari\u00e7) odakland\u0131k. \u00d6zellikle 1984 y\u0131l\u0131nda yay\u0131nlanan Fizik, Matematik ve Ekonomi alanlar\u0131n\u0131 ele ald\u0131k.<\/span><\/p>\n

Her alan i\u00e7in, yay\u0131nland\u0131ktan sonraki t y\u0131l boyunca K al\u0131nt\u0131 toplayan makalelerin oran\u0131, olas\u0131l\u0131k yo\u011funluk fonksiyonunu, p(K,t) verir. Kar\u015f\u0131l\u0131k gelen k\u00fcm\u00fclatif al\u0131nt\u0131 da\u011f\u0131l\u0131mlar\u0131n\u0131 g\u00f6sterir, P(K,t). Bu da\u011f\u0131l\u0131mlar\u0131 ayr\u0131k kuvvet yasas\u0131 ve ayr\u0131k log-normal fonksiyonu kullanarak yerle\u015ftirdik.<\/span><\/p>\n

\u0130lk uyum, log-log grafi\u011finde d\u00fcz bir kuyruk oldu\u011funu varsayarken, ikinci uyum, d\u0131\u015fb\u00fckey bir kuyruk varsayar. K\u00fc\u00e7\u00fck t i\u00e7in (yay\u0131nland\u0131ktan sonra erken) her iki uyumun da e\u015fit derecede iyi performans g\u00f6sterdi\u011fini, daha sonraki y\u0131llarda ise ayr\u0131k kuvvet yasas\u0131 uyumunun daha iyi oldu\u011funu g\u00f6sterir. Asl\u0131nda, Fizik makalelerinin en temsili k\u00fcmesi i\u00e7in, da\u011f\u0131l\u0131m\u0131n kuyru\u011fu \u00f6nerildi\u011fi gibi d\u0131\u015fb\u00fckey olmaktan ziyade d\u00fczd\u00fcr.<\/span><\/p>\n

\n\n
<\/div>\n\n<\/div>\n \n