Journal of Information and Organizational Sciences (JIOS)
Not a member yet
455 research outputs found
Sort by
SELECTING NEURAL NETWORK ARCHITECTURE FOR INVESTMENT PROFITABILITY PREDICTIONS
No full textAfter production and operations, finance and investments are one of the mostfrequent areas of neural network applications in business. The lack of standardizedparadigms that can determine the efficiency of certain NN architectures in a particularproblem domain is still present. The selection of NN architecture needs to take intoconsideration the type of the problem, the nature of the data in the model, as well as somestrategies based on result comparison. The paper describes previous research in that areaand suggests a forward strategy for selecting best NN algorithm and structure. Since thestrategy includes both parameter-based and variable-based testings, it can be used forselecting NN architectures as well as for extracting models. The backpropagation, radialbasis,modular, LVQ and probabilistic neural network algorithms were used on twoindependent sets: stock market and credit scoring data. The results show that neuralnetworks give better accuracy comparing to multiple regression and logistic regressionmodels. Since it is model-independant, the strategy can be used by researchers andprofessionals in other areas of application
Croatian banking sector research: relationship between ownership structure, concentration, owners’ type and bank performance
No full textBanks are important financial intermediaries of any national economy, and corporate governance has an important role in banking sector; especially due to processes of the globalization and the internationalization, and also because of the sensitivity of the activities between the interest groups. The objective of this paper is to examine the relationship between ownership structure, concentration, owners’ type and bank performance. The authors made a research of banks' ownership structure using publicly available data. Using statistical tools authors discovered relationships between bank ownership structure and bank performance indicators (average asset, total asset, average equity, profit (loss) before taxes, profit (loss) after taxes, ROAA, ROAE). Further they discuss the relationships between ownership structure and a number of consequences for the bank performance. The authors discovered significant correlation between bank ownership structure and performance indicators variables that are described in the paper
From the Editor
No full textDear authors, readers and future authors,It is my pleasure to present you the 2nd number of the 36th volume of the Journal of Information andOrganizational Sciences – JIOS.In this issue a selection of 3 original scientic papers, 3 preliminary communications and 1 surveypaper has been included, all of which have undergone a rigorous double-blind review process insome cases in several rounds.As in prior issues of the Journal, this one also comprises papers that cover themes from various ICTelds and related areas: computer algorithms, ERP systems, business process analysis, and otherrelated themes. We hope that the variety of themes will draw the attention of researchers in dierentelds of information sciences and motivate potential authors to expand the Journal’s thematic scopeby other themes that may be of interest to the wider ICT research community. These themes motivatesus to continue our work of delivering research results to interested audience. We hope that many ofyou who will read the Journal will join us as authors, readers and reviewers and help us to fulll ourmission. The number of submitted papers increases every year giving us opportunity to increase thequality and visibility of JIOS.I would like to use this opportunity to encourage prospective authors among you, dear readers, tosubmit papers at http://jios.foi.hr. You are also invited to share with us your comments or criticalremarks regarding the papers previously published in the Journal. Further, we would appreciate yourwillingness to become a reviewer for submitted papers and thus contribute signicantly to the risingquality of JIOS.So far JIOS has been indexed in nine well known databases. We are continuously investing a lot ofeort to extend the list of databases where JIOS would be cited, in order to make the publishedpapers available to a larger readers’ audience, and give your research a greater impact factor
SIGMA-NOTATION AND THE EQUIVALENCE OF P AND NP CLASSES
No full textThe aim of this paper is to show that –notation, based on the positionalityprinciple, doesn’t just answer the question of equivalence of P and NP problemclasses, but also represents the key to the solution of (some) more general problemsfrom the domain of multivalued logics. However, the achievement of such an aim inits full scope is hardly possible within the limits of this paper. Therefore, the followingplan shall be realized:Firstly, the logic algebra fragment necessary for the solution of the P and NPclasses equivalence problem shall be expounded.Secondly, the necessary symbols and definitions shall be introduced to show thatthe calculations in the domain of the multivalued logic can be directly executed withinthe framework of –notation
WEB LOG EXPLORER – CONTROL OF MULTIDIMENSIONAL DYNAMICS OF WEB PAGES
No full textDemand markets dictate and pose increasingly more requirements to the supplymarket that are not easily satisfied. The supply market presenting its web pages to thedemand market should find the best and quickest ways to respond promptly to the changesdictated by the demand market. The question is how to do that in the most efficient andquickest way. The data on the usage of web pages on a specific web site are recorded in alog file. The data in a log file are stochastic and unordered and require systematicmonitoring, categorization, analyses, and weighing. From the data processed in this way, itis necessary to single out and sort the data by their importance that would be a basis for acontinuous generation of dynamics/changes to the web site pages in line with the criterionchosen. To perform those tasks successfully, a new software solution is required. For thatpurpose, the authors have developed the first version of the WLE (WebLogExplorer)software solution, which is actually realization of web page multidimensionality and theweb site as a whole. The WebLogExplorer enables statistical and semantic analysis of a logfile and on the basis thereof, multidimensional control of the web page dynamics. Theexperimental part of the work was done within the web site of HTZ (Croatian NationalTourist Board) being the main portal of the global tourist supply in the Republic of Croatia(on average, daily "log" consists of c. 600,000 sets, average size of log file is 127 Mb, andc. 7000-8000 daily visitors on the web site)
THE RICH WORLD OF COMPUTER MODELLING AND PROBLEM SOLVING
No full textThis paper presents some interesting aspects of the rich world of computermodelling and problem solving. After discussing basic model building principles attentionis devoted to several modelling methods aimed for solving different types of problems.These are methods for modelling of optimisation problems, dynamic systems with randomcharacteristics, knowledge, uncertain phenomena, systems with graph structure, learning,as well as cooperation and competition. Basic characteristics of these methods andexamples of their application are given
LOG FILE ANALYSIS AND CREATION OF MORE INTELLIGENT WEB SITES
No full textTo enable successful performance of any company or business system, both inthe world and in the Republic of Croatia, among many problems relating to its operationsand particularly to maximum utilization and efficiency of the Internet as a media forrunning business (especially in terms of marketing), they should make the best possible useof the present-day global trends and advantages of sophisticated technologies andapproaches to running a business. Bearing in mind the fact of daily increasing competitionand more demanding market, this paper addresses certain scientific and practicalcontribution to continuous analysis of demand market and adaptation thereto by analyzingthe log files and by retroactive effect on the web site. A log file is a carrier of numerousdata and indicators that should be used in the best possible way to improve the entirebusiness operations of a company. However, this is not always simple and easy. The websites differ in size, purpose, and technology used for designing them. For this very reason,the analytic analysis frameworks should be such that can cover any web site and at thesame time leave some space for analyzing and investigating the specific characteristicof each web site and provide for its dynamics by analyzing the log file records. Thoseconsiderations were a basis for this pape
INDEPENDENT DE-DUPLICATION IN DATA CLEANING
No full textMany organizations collect large amounts of data to support their business anddecision-making processes. The data originate from a variety of sources that may haveinherent data-quality problems. These problems become more pronounced whenheterogeneous data sources are integrated (for example, in data warehouses). A majorproblem that arises from integrating different databases is the existence of duplicates. Thechallenge of de-duplication is identifying “equivalent” records within the database. Mostpublished research in de-duplication propose techniques that rely heavily on domainknowledge. A few others propose solutions that are partially domain-independent. Thispaper identifies two levels of domain-independence in de-duplication namely: domainindependenceat the attribute level, and domain-independence at the record level. Thepaper then proposes a positional algorithm that achieves domain-independent deduplicationat the attribute level, and a technique for field weighting by data profiling,which, when used with the positional algorithm, achieves domain-independence at therecord level. Experiments show that the proposed techniques achieve more accurate deduplicationthan the existing algorithms
