https://phaidrabg.bg.ac.rs/o:30204 srp Sadržaj – Tema ovog rada je proučavanje dinamike i statističkih osobina Knjige graničnih naloga. Da bi se opisala složena dinamika trgovanja prisutna na tržištu, koriste se koncepti mašinskog učenja. Veštačka inteligencija postaje sve više prisutna u raznim oblastima. Potencijalna primenu koncepta mašinskog učenja u finansijama zainteresovala je veliki broj istrazivaca kako iz akademije tako i iz industrije. Pošto je trgovanje doživelo automatizaciju i digitalizaciju, koncept fizičkog trgovanja je zamenjen elektronskim trgovanjem. Zbog toga postoji veliko interesovanje u razvijanju algoritama za automatsko trgovanje zasnovanih na mašinskom učenju. Cilj ovog istraživanja je da proučava distribuciju varijable koja modelira razliku između najmanje cene na strani za kupovinu, i najvece cene na strani za prodaju. S obzirom da su finansijska tržišta informativna, postoji potencijal u analizi istorijskih podataka o akcijama, kao i u razvoju algoritama za analizu tih podataka. Abstract - The topic of this paper is the study of the dynamics and statistical properties of the Limit Order Book. In order to describe the complex trading dynamics present in the market, machine learning concepts are used. Artificial intelligence is becoming more and more present in various fields. The potential application of the machine learning concepts in finance has interested a large number of researchers from both academia and industry. As trading has undergone automation and digitalization, the concept of physical trading has been replaced by electronic trading. Therefore, there is great interest in developing machine learning algorithms based on machine learning. The aim of this study is to study the distribution of a variable that models the difference between the lowest price on the buying side and the highest price on the selling side. Since financial markets are informative, there is potential in analyzing historical stock data, as well as in developing algorithms for analyzing that data. info:eu-repo/semantics/conferenceProceedings Radojičić, Dragana All rights reserved 2022 application/pdf 1059226 bytes Zbornik radova 28. IKT konferencije “YU INFO 2022” Primena masinskog učenja u modeliranju dinamike knjige limitiranih naloga Application of machine learning in modeling the dynamics of the limit order