20.500.12556/DKUM-84824 Primerjava uspešnosti napovedi stečaja podjetja tradicionalnega Z-modela z novejšim Ycz-modelom Comparing the performance of traditional Z- and newer Ycz- bankruptcy prediction models V diplomskem delu obravnavamo področje napovedi stečaja podjetja, pri čemer tudi primerjamo uspešnost tradicionalnega Zmijewskijevega Z-modela z novejšim Ycz-modelom. Napoved stečaja je zelo koristna za mnoge deležnike, kot so upniki, investitorji, poslovni partnerji, dobavitelji in kupci. Da bi videli, kako zdravo je poslovanje podjetja, se ves čas razvijajo napovedni modeli. Ti modeli podjetjem omogočajo, da bolje zaščitijo svoje poslovanje, sprejmejo določene ukrepe in celo preprečijo stečaj. Modele oziroma metode na splošno delimo v dve skupini, in sicer ločimo statistične metode in metode strojnega učenja. Izbrana modela temeljita na logitni regresiji in multivariatni diskriminantni analizi, ki obe spadata pod statistične metode. V delu ugotavljamo, kako uspešna sta izbrana modela za napoved stečaja slovenskih podjetij eno, dve in tri leta pred začetkom stečajnega postopka. Analiza je izvedena na treh poljubno izbranih propadlih slovenskih podjetjih, ki so šla v stečaj v zadnjih petih letih. Na podlagi izvedene analize ugotavljamo, da je Ycz-model uspešnejši, saj večkrat pravilno in ustrezno napove stečaj. In the bachelor’s thesis, we discuss the field of company bankruptcy forecasting while also comparing the performance of the traditional Zmijewski Z-model with the newer Ycz model. A declaration of bankruptcy is very beneficial for many stakeholders, such as creditors, investors, business partners, suppliers, and customers. With the purpose to realize how healthy a company's business is, predictive models are constantly being developed. This allows companies to protect their operations better, take certain measures, and even prevent bankruptcy. We generally divide models or methods into two groups: statistical methods and machine learning methods. The selected models are based on logit regression and multivariate discriminant analysis both of which belong to statistical methods. In the thesis, we ascertain how successful the selected models are for predicting the bankruptcy of Slovenian companies one, two, and three years before the start of bankruptcy proceedings. The analysis was carried out on three arbitrarily selected failed Slovenian companies that went bankrupt in the last five years. Based on the performed analysis, we conclude that the Ycz model is more successful because it predicts bankruptcy correctly and appropriately several times. stečaj podjetja modeli napovedi stečaja podjetja Zmijewskijev model Ycz-model propadla slovenska podjetja company bankruptcy company bankruptcy prediction models Zmijewski's model Ycz-model bankrupted Slovenian companies true true false V. Černesl Slovenski jezik Angleški jezik Diplomsko delo/naloga 2023-07-28 22:32:47 2023-09-14 11:37:27 2023-09-15 03:06:54 0000-00-00 00:00:00 2023 0 Ptuj 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 005.521 164426243 UN_Cernesl_Vida_2023.pdf UN_Cernesl_Vida_2023.pdf 1 096525A2D7FBCFCDDD0CE8B7D7537725 5454fdb9a8f5856be28e453bb54d59804809d9ccc8bbed1f4286973fd463888f f6a0f7b3-3d3f-11ee-af21-00155d000105 https://dk.um.si/Dokument.php?lang=slv&id=169715 Ekonomsko-poslovna fakulteta 0 0 0