1. Modeli slučajnih šuma i primjene
- Author
-
Čular, Marko and Slijepčević, Siniša
- Subjects
machine learning ,slučajna šuma ,PRIRODNE ZNANOSTI. Matematika ,decision tree ,linear regression ,linearna regresija ,NATURAL SCIENCES. Mathematics ,random forest ,strojno učenje ,stablo odluke - Abstract
U ovome radu glavni predmet proučavanja bila je primjena modela slučajnih šuma u svakodnevnom života, točnije u računanju kreditnog rizika u bankama. U početnom dijelu rada pobliže smo upoznali čitatelja s jednostavnom i višestrukom linearnom regresijom. U nastavku rada donijeli smo pregled razvoja strojnog učenja te pojasnili istoimeni pojam. Zatim smo detaljno opisali pojam stabla odluke i kako se ono razvija te smo njegov model prikazali na primjeru i vizualno radi lakšeg shvaćanja napisanog. Slijedilo je poglavlje posvećeno teorijskoj obradi modela slučajnih šuma i njihovoj primjeni u bankarstvu. Radi lakše interpretacije podaci su prikazani kroz histograme i tablice. Za kraj rada donesi su zaključci o radu algoritma i mogućnosti njegove implementacije u bankarski sustav. In this paper, the main subject of study was application of the model of random forests in everyday life, more precisely in the calculation of credit risk in banks. In the initial part of the paper, we introduced the reader to simple and multiple linear regression. In the continuation of the paper, we provide an overview of the development of machine learning and clarify the concept of them. Then we minutely described the concept of the decision tree and how it develops, and we presented its model on an example and visually to make it easier to understand what is written. This was followed by a chapter dedicated to the theoretical processing of random forest models and their application in banking. For easier interpretation, the data are presented through histograms and tables. At the end of the paper, conclusions were made about the operation of the algorithm and the possibility of its implementation in the banking system.
- Published
- 2020