Jedan od osnovnih zadataka statistike je procjena raspodjele posmatranog obilježja populacije. U klasičnom parametarskom pristupu, prvo se odlučujemo za tip raspodjele (tj. statistički model), a zatim određenim metodama procjenjujemo nepoznate parametre modela na osnovu uzorka. Međutim, ponekad želimo isprobati više konkurentskih modela i odabrati onaj koji obezbjeđuje najbolje fitovanje. U radu prikazujemo neke goodness-of-fit statistike (Kolmogorov-Smirnov, Anderson-Darling, Cramervon Mises), kao i informacijske kriterije (Akaike, Bayes), koji mogu pomoći u izboru adekvatnog modela. Kriterije ćemo implementirati u R-u, a ilustrovati na simuliranim, kao i stvarnim podacima.
Cite as:
Safet Hamedović, Hermina Alajbegović, Almir Huskanović (2025) Izbor adekvatnog statističkog modela. Quality 2025 (S. Jašarević, editor), ISSN 1512-9268, Neum, B&H, 3-5 June 2025., pp. 49-54
DISCLAIMER: Professional – scientific papers published in Conference proceedings are published in the original. Papers were reviewed by members of the Scientific Committee. Authors are responsible for technical and linguistic correctness of the text. WARNING: This is an open access publication which means that all content is freely available without charge to the user or his/her institution. Users are allowed to read, download, copy, distribute, print, search, or link to the full texts of the articles in this journal without asking prior permission from the publisher or the author. This is in accordance with the BOAI definition of open access. The greatest possible effort is put to ensure that the Proceedings are complete and accurate, but it does not imply any warranty or liability. The authors and publisher have no obligation or responsibility to any person or entity in connection with any damage or loss due to the information published in the Conference Proceedings.
