{"id":12268,"date":"2023-05-23T13:37:30","date_gmt":"2023-05-23T10:37:30","guid":{"rendered":"https:\/\/bestessayhomework.com\/tr\/?p=12268"},"modified":"2023-05-23T13:37:30","modified_gmt":"2023-05-23T10:37:30","slug":"atif-dinamikleri","status":"publish","type":"post","link":"https:\/\/bestessayhomework.com\/tr\/atif-dinamikleri\/","title":{"rendered":"At\u0131f Dinamikleri"},"content":{"rendered":"

At\u0131f Dinamikleri<\/span><\/strong><\/h3>\n

\u00d6l\u00e7\u00fclen ve sim\u00fcle edilen at\u0131f da\u011f\u0131l\u0131mlar\u0131 aras\u0131ndaki uyum etkileyici olsa da, bu ger\u00e7ek tek ba\u015f\u0131na modeli do\u011frulamak i\u00e7in yeterli de\u011fildir. Ayn\u0131 da\u011f\u0131l\u0131mlar \u00e7ok farkl\u0131 at\u0131f yollar\u0131ndan kaynaklanabilir. \u00d6rne\u011fin, sars\u0131nt\u0131l\u0131 ve p\u00fcr\u00fczs\u00fcz at\u0131f y\u00f6r\u00fcngeleri, ayn\u0131 at\u0131f da\u011f\u0131l\u0131mlar\u0131 setini verebilir.<\/span><\/p>\n

Bu nedenle, ek bir test ger\u00e7ekle\u015ftirdik ve modelin yaln\u0131zca farkl\u0131 zamanlarda at\u0131f da\u011f\u0131l\u0131mlar\u0131n\u0131 do\u011fru bir \u015fekilde yeniden \u00fcretti\u011fini de\u011fil, ayn\u0131 zamanda tek tek makalelerin at\u0131f y\u00f6r\u00fcngelerini de \u00fcretti\u011fini do\u011frulad\u0131k.<\/span><\/p>\n

Modelin, benzer makale gruplar\u0131 i\u00e7in \u00f6l\u00e7\u00fclen ve sim\u00fcle edilmi\u015f at\u0131f yollar\u0131n\u0131n da\u011f\u0131l\u0131mlar\u0131n\u0131 do\u011fru bir \u015fekilde yeniden \u00fcretti\u011fini g\u00f6sterin. A\u015fa\u011f\u0131da, modelin dalgalanmalar\u0131, yani al\u0131nt\u0131 y\u00f6r\u00fcngelerinin sars\u0131nt\u0131s\u0131n\u0131 ne kadar iyi yeniden \u00fcretti\u011fini analiz edece\u011fiz.<\/span><\/p>\n

Makalelerin al\u0131nt\u0131 dinamiklerinin dalgalanan bile\u015fenini \u00f6l\u00e7mek i\u00e7in, al\u0131nt\u0131lar ayr\u0131k oldu\u011fundan ve al\u0131nt\u0131 y\u00f6r\u00fcngesinin bu iki bile\u015feni birbirinden ayr\u0131lamaz oldu\u011fundan, makalelerin al\u0131nt\u0131 y\u00f6r\u00fcngelerini d\u00fczg\u00fcn ve dalgalanan bile\u015fenlere ayr\u0131\u015ft\u0131rmak amac\u0131yla analiz etmiyoruz.<\/span><\/p>\n

Ayr\u0131ca, at\u0131f dinami\u011fi zamana ba\u011fl\u0131 homojen olmayan stokastik bir s\u00fcre\u00e7 oldu\u011fundan, bir makalenin at\u0131f y\u00f6r\u00fcngesinin yumu\u015fat\u0131lmas\u0131 iyi tan\u0131mlanmam\u0131\u015ft\u0131r. Bu zorluklar\u0131n \u00fcstesinden gelmek i\u00e7in farkl\u0131 bir yakla\u015f\u0131m benimsedik ve benzer makale setlerinin at\u0131f dinamiklerinin istatistiksel karakterizasyonunu ama\u00e7lad\u0131k.<\/span><\/p>\n

Ger\u00e7ekten de, modelimiz \u00e7er\u00e7evesinde, bir makalenin at\u0131f dinamiklerini belirleyen temel bireysel parametre, uygunlu\u011fudur. Ayn\u0131 y\u0131lda yay\u0131nlanan ve hepsi ayn\u0131 uygunlukta olan bir disiplinde bir dizi ara\u015ft\u0131rma makalesi i\u00e7in, sabit bir zamanda al\u0131nt\u0131lar\u0131n istatistiklerini analiz ettik.<\/span><\/p>\n

Bu k\u00fcmenin al\u0131nt\u0131 y\u00f6r\u00fcngelerinin bir dizi anl\u0131k g\u00f6r\u00fcnt\u00fcs\u00fcn\u00fc kullanarak, ek al\u0131nt\u0131lar\u0131n istatistiksel \u00f6zelliklerini \u00f6l\u00e7t\u00fck ve ortalamalar\u0131n\u0131, varyanslar\u0131n\u0131 ve otokorelasyonlar\u0131n\u0131 bulduk. Varyans, dalgalanmalar\u0131n b\u00fcy\u00fckl\u00fc\u011f\u00fcn\u00fcn bir \u00f6l\u00e7\u00fcs\u00fcd\u00fcr, otokorelasyon katsay\u0131s\u0131 ise s\u00fcrelerini \u00f6l\u00e7er.<\/span><\/p>\n

Buradaki zorluk, ayn\u0131 uygunluktaki ka\u011f\u0131tlar\u0131n nas\u0131l bulunaca\u011f\u0131d\u0131r. Bu ama\u00e7la, her makalenin al\u0131nt\u0131 y\u00f6r\u00fcngesini analiz etmemiz ve bunu modelle kar\u015f\u0131la\u015ft\u0131rmam\u0131z gerekiyordu. K\u0131sa yoldan gittik ve ayn\u0131 y\u0131l i\u00e7inde yay\u0131nlanan ve belirli bir t s\u00fcresine kadar ayn\u0131 say\u0131da at\u0131f alan bir alandaki makalelerin ayn\u0131 uygunlukta oldu\u011funu varsayd\u0131k.<\/span><\/p>\n

Ek At\u0131flar\u0131n \u0130statistiksel Da\u011f\u0131l\u0131m\u0131<\/span><\/strong><\/h3>\n

At\u0131f dinamiklerinin p\u00fcr\u00fczs\u00fcz k\u0131sm\u0131n\u0131 karakterize etmek i\u00e7in, ayn\u0131 makale setini kulland\u0131k, yani 1984’te yay\u0131nlanan Fizik makalelerini, yay\u0131ndan sonraki t y\u0131l boyunca toplanan al\u0131nt\u0131lar\u0131n say\u0131s\u0131 olan ayn\u0131 K(t)’ye sahip makaleleri i\u00e7eren kutulara ay\u0131rd\u0131k.\u00a0<\/span><\/p>\n

Her b\u00f6lmedeki makaleler i\u00e7in, ek al\u0131nt\u0131lar\u0131n istatistiksel da\u011f\u0131l\u0131m\u0131n\u0131 dikkate ald\u0131k, kj = \u0394Kj (t + 1 y\u0131l\u0131nda toplanan al\u0131nt\u0131lar). Bu, \u00f6l\u00e7\u00fclen ve sim\u00fcle edilen veriler i\u00e7in de yap\u0131ld\u0131. \u00d6l\u00e7\u00fclen ve sim\u00fcle edilen da\u011f\u0131l\u0131mlar i\u00e7in ortalama ek al\u0131nt\u0131 say\u0131s\u0131n\u0131n (kj) \u00e7ok yak\u0131n oldu\u011funu g\u00f6stermektedir. Bundan, modelin makalelerin al\u0131nt\u0131 dinamiklerinin p\u00fcr\u00fczs\u00fcz k\u0131sm\u0131n\u0131 olduk\u00e7a iyi yakalad\u0131\u011f\u0131 sonucuna vard\u0131k.<\/span><\/p>\n

\n

Makalede at\u0131f<\/a> nedir<\/span>
\nAt\u0131f \u00d6rnekleri<\/span>
\nMakalede at\u0131f nas\u0131l yap\u0131l\u0131r<\/span>
\nBilimsel at\u0131f nedir<\/span>
\nAt\u0131fta bulunmak \u00f6rnek<\/span>
\nAl\u0131nt\u0131 ve at\u0131f aras\u0131ndaki fark<\/span>
\nAt\u0131f nas\u0131l yap\u0131l\u0131r<\/span>
\nAt\u0131f Ne Demek \u00f6rnek<\/span><\/p>\n

\n

Otokorelasyon<\/span><\/strong><\/h3>\n

At\u0131f dalgalanmalar\u0131n\u0131n s\u00fcresini karakterize etmek i\u00e7in, sonraki y\u0131llarda bir makale taraf\u0131ndan elde edilen ek at\u0131flar aras\u0131ndaki otokorelasyonlar\u0131 dikkate ald\u0131k. \u00d6nceki at\u0131f say\u0131s\u0131 K(t) ile ayn\u0131 olan makaleler i\u00e7in \u00f6l\u00e7\u00fclen y\u0131ll\u0131k at\u0131flar i\u00e7in hesaplad\u0131k. Spesifik olarak, kj (t ) ve kj (t \u2212 1) olmak \u00fczere birbirini takip eden iki y\u0131l boyunca her makalenin ald\u0131\u011f\u0131 al\u0131nt\u0131 say\u0131s\u0131n\u0131 belirledik ve bu k\u00fcme i\u00e7in Pearson otokorelasyon katsay\u0131s\u0131n\u0131 hesaplad\u0131k.<\/span><\/p>\n

kj (t) ve kj (t \u2212 1), s\u0131ras\u0131yla kj (t) ve kj (t \u2212 1) da\u011f\u0131l\u0131mlar\u0131n\u0131n ortalamas\u0131 ve \u03c3 (t ) ve \u03c3 (t \u2212 1) standart sapmalard\u0131r.<\/span><\/p>\n

ct,t-1’in fiziksel anlam\u0131 nedir? At\u0131flar Markov s\u00fcrecini takip ediyorsa, bir makalenin at\u0131f oran\u0131, birikmi\u015f at\u0131flar\u0131n toplam say\u0131s\u0131na g\u00f6re ve \u00f6nceki at\u0131f ge\u00e7mi\u015finden ba\u011f\u0131ms\u0131z olarak ct,t\u22121 = 0 olacak \u015fekilde belirlenmelidir. \u00d6te yandan, ct ,t\u22121 = 1, bir makalenin at\u0131f oran\u0131n\u0131n tamamen ge\u00e7en y\u0131l toplanan at\u0131f say\u0131s\u0131na g\u00f6re belirlendi\u011fini g\u00f6sterir.<\/span><\/p>\n

B\u00f6ylece, ct,t-1, al\u0131nt\u0131 s\u00fcrecinin Markoviyenden sapmas\u0131n\u0131 \u00f6l\u00e7er. Ba\u015fka bir a\u00e7\u0131dan ct,t\u22121 = 0, sonraki y\u0131llarda at\u0131f oranlar\u0131n\u0131n ili\u015fkisiz oldu\u011funu, yani dalgalanmalar\u0131n karakteristik s\u00fcresinin 1 y\u0131ldan az oldu\u011funu g\u00f6sterir. \u00d6te yandan, ct,t-1 = 1, bir makalenin sonraki y\u0131llarda at\u0131f oranlar\u0131 aras\u0131nda g\u00fc\u00e7l\u00fc bir pozitif korelasyon oldu\u011funu g\u00f6sterir.<\/span><\/p>\n

\u015eekil 5.6, ct,t-1’in t\u00fcm makaleler i\u00e7in ayn\u0131 olmad\u0131\u011f\u0131n\u0131, ancak birikmi\u015f at\u0131f say\u0131s\u0131 K ile b\u00fcy\u00fcd\u00fc\u011f\u00fcn\u00fc g\u00f6stermektedir. D\u00fc\u015f\u00fck at\u0131fl\u0131 makaleler i\u00e7in ct ,t \u22121 k\u00fc\u00e7\u00fckt\u00fcr. Bu, bu makalelerin al\u0131nt\u0131 dinamiklerinin bir Markov s\u00fcreci ile a\u00e7\u0131klanabilece\u011fini g\u00f6sterir.<\/span><\/p>\n

Bu, ct,t\u22121 \u223c 1 olan y\u00fcksek oranda at\u0131f yap\u0131lan makalelerin tersidir ve bu makalelerin at\u0131f dinamiklerinin, at\u0131f ge\u00e7mi\u015flerinden b\u00fcy\u00fck \u00f6l\u00e7\u00fcde etkilendi\u011fini ve bir Markov s\u00fcrecini takip etmedi\u011fini g\u00f6sterir. Bu d\u00fc\u015f\u00fcnceler ayn\u0131 zamanda, y\u00fcksek oranda at\u0131f alan makalelerin d\u00fczg\u00fcn gidi\u015fatlar\u0131n\u0131n aksine, d\u00fc\u015f\u00fck at\u0131f alan makalelerin sars\u0131nt\u0131l\u0131 at\u0131f y\u00f6r\u00fcngelerini de a\u00e7\u0131klar.<\/span><\/p>\n

Modelimiz \u00f6l\u00e7\u00fclen ct,t\u22121(K) ba\u011f\u0131ml\u0131l\u0131klar\u0131n\u0131 t > 10 i\u00e7in olduk\u00e7a iyi bir \u015fekilde yeniden \u00fcretir, ancak t < 10 i\u00e7in sim\u00fcle edilmi\u015f ct,t\u22121(K) de\u011ferleri \u00f6l\u00e7\u00fclen de\u011ferlerden daha k\u00fc\u00e7\u00fckt\u00fcr.<\/span><\/p>\n

D\u00fc\u015f\u00fck ve y\u00fcksek at\u0131fl\u0131 makalelerin at\u0131f dinamikleri aras\u0131ndaki \u00e7arp\u0131c\u0131 fark (Markovyan ve Markovyan olmayan s\u00fcre\u00e7), P0 (K ) ba\u011f\u0131ml\u0131l\u0131\u011f\u0131ndan kaynaklanan do\u011frusal olmaman\u0131n do\u011frudan bir sonucudur. G\u00f6sterilen ct,t\u22121(K) ba\u011f\u0131ml\u0131l\u0131\u011f\u0131n\u0131n ayn\u0131 kaynaktan geldi\u011fini iddia ediyoruz. Bunu g\u00f6stermek i\u00e7in, modelin s\u00fcrekli yakla\u015f\u0131m\u0131na gidiyoruz ve 1\/\u03b3 \u0303 \u2248 0.8’in 1 y\u0131la yak\u0131n oldu\u011fu i\u00e7in bu denklemin yeniden d\u00fczenlenebilece\u011fini fark ediyoruz.<\/span><\/p>\n

Asl\u0131nda, farkl\u0131 y\u0131llar i\u00e7in ct ,t \u22121 ba\u011f\u0131ml\u0131l\u0131klar\u0131n\u0131n tek bir e\u011fri \u00fczerine \u00e7\u00f6kt\u00fc\u011f\u00fcn\u00fc g\u00f6sterir. Dolay\u0131s\u0131yla otokorelasyon katsay\u0131s\u0131, dolayl\u0131 al\u0131nt\u0131lar\u0131n olas\u0131l\u0131\u011f\u0131 P0 ile do\u011frudan ili\u015fkili olan q parametresinin do\u011frudan bir \u00f6l\u00e7\u00fcs\u00fcd\u00fcr. Asl\u0131nda, ct,t\u22121(K) ba\u011f\u0131ml\u0131l\u0131\u011f\u0131n\u0131n q(K) ba\u011f\u0131ml\u0131l\u0131\u011f\u0131n\u0131 taklit etti\u011fini g\u00f6sterir.<\/span><\/p>\n

Makalelerin at\u0131f dinamiklerinin dalgalanan k\u0131sm\u0131n\u0131 \u00f6l\u00e7mek i\u00e7in, benzer makaleler i\u00e7in, yani ayn\u0131 zaman periyodunda ayn\u0131 say\u0131da K(t) al\u0131nt\u0131 toplayanlar i\u00e7in ek at\u0131flar\u0131n (kj) istatistiksel da\u011f\u0131l\u0131m\u0131n\u0131n varyans\u0131n\u0131 dikkate ald\u0131k.<\/span><\/p>\n

Bu varyans iki kaynaktan gelir: (1) at\u0131f s\u00fcrecinin olas\u0131l\u0131ksal karakterinden kaynaklanan i\u00e7sel varyans ve (2) bu k\u00fcmedeki makalelerin farkl\u0131 at\u0131f ge\u00e7mi\u015fleriyle ili\u015fkili at\u0131f dinamiklerinin de\u011fi\u015fkenli\u011fi. \u0130kinci de\u011fi\u015fkenlik kayna\u011f\u0131, bir makalenin at\u0131f dinami\u011finin bir Markov s\u00fcreci olmamas\u0131ndan kaynaklanmaktad\u0131r, yaln\u0131zca birikmi\u015f at\u0131f say\u0131s\u0131na (K) de\u011fil, bir makalenin t\u00fcm at\u0131f ge\u00e7mi\u015fine de ba\u011fl\u0131d\u0131r.<\/span><\/p>\n

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