Napovedovanje verjetnosti neplačila  z nevronskimi mrežami

Rajter, Urban

| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

First page > Show document

Show document

Title:	Napovedovanje verjetnosti neplačila z nevronskimi mrežami
Authors:	ID Rajter, Urban (Author) ID Taranenko, Andrej (Mentor) More about this mentor... ID Stanet, Peter (Comentor)
Files:	MAG_Rajter_Urban_2021.pdf (1,95 MB) MD5: 7E8465B6629C70A33FC80968B8CD836E PID: 20.500.12556/dkum/b8b27552-5d56-46fc-a7da-9f38d4b4a584
Language:	Slovenian
Work type:	Master's thesis/paper
Typology:	2.09 - Master's Thesis
Organization:	FNM - Faculty of Natural Sciences and Mathematics
Abstract:	Umetna inteligenca se nanaša na teorijo in razvoj računalniških sistemov, ki lahko opravljajo naloge, ki običajno zahtevajo človeško inteligenco. Podskupina strojnega učenja je globoko učenje, kjer se umetne nevronske mreže, algoritmi, ki jih navdihujejo človeški možgani, učijo iz velikih količin podatkov. Podobno, kot se ljudje učimo iz izkušenj, bi algoritem globokega učenja večkrat ponovil isto nalogo in jo vsakič nekoliko prilagodil, da bi izboljšal rezultat. V tej magistrski nalogi so predstavljene nevronske mreže, tipi nevronskih mrež in njihova uporaba. Podrobneje je opisana uporaba nevronskih mrež za namene napovedovanja verjetnosti neplačila. Uporabljen je model globoke nevronske mreže na anonimiziranih podatkih podjetja. Opisan je postopek priprave podatkov in postopek učenja modela na vhodnih podatkih. Analiza končnega rezultata pove, da je uporaba nevronskih mrež smiselna, saj algoritem nudi visoko natančnost.
Keywords:	strojno učenje, nevronske mreže, globoko učenje, globoke nevronske mreže, kreditno tveganje
Place of publishing:	Maribor
Publisher:	[U. Rajter]
Year of publishing:	2021
PID:	20.500.12556/DKUM-79310
UDC:	004.85:519.22(043.2)
COBISS.SI-ID:	70537219
Publication date in DKUM:	02.08.2021
Views:	1613
Downloads:	187
Metadata:
Categories:	FNM
:	Copy citation

Average score:	(0 votes)
Your score:	Voting is allowed only for logged in users.
Share:

Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Licences

License:	CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Link:	http://creativecommons.org/licenses/by-nc-nd/4.0/
Description:	The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.
Licensing start date:	10.06.2021

Secondary language

Language:	English
Title:	Forecasting probability of default with neural networks
Abstract:	Artificial intelligence refers to the theory and development of computer systems that can perform tasks, which typically require human intelligence. Deep learning is a subset of machine learning where artificial neural networks, algorithms inspired by the human brain, learn from large amounts of data. Similarly to how people learn from experience, the deep learning algorithm would repeat the same task several times and each time adjust it slightly in order to improve the result. In this master’s thesis, we present neural networks, types of neural networks and their usage. We also thoroughly describe the use of neural networks for the purpose of forecasting the probability of default. A deep neural network algorithm is used on anonymized company data. The whole process includes data preparation and teaching the algorithm on the input data. The final analysis shows that neural networks are suitable for our problem, because the algorithm provides high accuracy.
Keywords:	machine learning, neural networks, deep learning, deep neural networks, credit risk

Comments

Leave comment

You must log in to leave a comment.

Comments (0)

0 - 0 / 0

There are no comments!

Back