Search

Credit risk assessment process includes evaluation of loan applications (approval of acceptable clients and rejection of clients that are likely to default) using application models, as well as monitoring behavior of existing clients using behavioral models. In this article we propose a deep self-normalizing neural network behavioral model trained on a large contract-level dataset. The proposed deep learning model outperformed conventional logistic regression based methods, with out-of-sample Somers’ D score of 84.08%. Moreover, when comparing accuracy scores with regard to actual month of default in the future, deep model once again exhibits higher predictive power.

True physical quantum computer is not jet widely available to researchers, so simulation plays an important role in understanding existent and development of new quantum algorithms and circuitry. This paper discusses simulation of quantum Shor&#8217 ; s algorithm for integer factorization. Simple simulation tool is presented. Several simulation techniques are discussed.

Minijaturizacijom elektroničkih komponenti fizikalni proces obrade informacije postaje dominantno kvantno-mehaničke prirode. Ispoljavanje kvantnih efekata u ponašanju osnovnih elemenata zahtijeva promjenu načina razmišljanja u teoriji informacije i računarstvu. Ovaj rad uvodi model kvantnog računala te na njemu analizira kvantnu prirodu obrade podataka. Uvodi se pojam kvantnog algoritma te se analiziraju prednosti kvantnog računala nad klasičnim. Posebno se analizira Shorov algoritam za faktorizaciju prirodnih brojeva. Kako stvarna kvantna računala još nisu široko dostupna, poseban se naglasak stavlja na problematiku simulacije kvantnih računala i procesa uopće. Predstavljen je i jednostavan simulacijski program.