No. 3/2016

Online archive of Computer Science and Mathematical Modelling

No. 3/2016

  1. K. Antczak - Rank thresholds in classifier ensemles in medical diagnosis
    Pages: 5 - 12
    Abstract: Classification methods have multiple applications, with medical diagnosis being one of the most common. A powerful way to improve classification quality is to combine single classifiers into an ensemble. One of the approaches for creating such ensembles is to combine class rankings from base classifiers. In this paper, two rank-based ensemble methods are studied: Highest Rank and Borda Count. Furthermore, the effect of applying class rank threshold to these methods is analyzed. We performed tests using real-life medical data. It turns out that specificity of data domain can affect classification quality depending on classifier type.
    Keywords medical diagnostics, classifier ensemble, rank threshold
    Full article: PDF icon 5_12_csmm3.pdf
  2. A. Chojnacki, G. Pieniążek - Choosing the optimal strategy for information security in a business organization
    Pages: 13 - 22
    Abstract: The paper describes the method of choosing the optimal strategy to implement security measures in a business organization. Strategies are categorized depending on time horizons, the history of threats and implemented security measures. Next, the method of choosing the optimal strategy for a business organization in a given context is outlined. Then this method is used to select the optimal strategy in a particular business context. The method is based on a deterministic time-based information security model, which was extended to a random model. With this simulation method, an organization can choose a strategy to implement security measures that best suits its needs. It is important for organizations to conduct an analysis of costs and threats in order to select appropriate safeguards.
    Keywords information security, stochastic model, simulation
    Full article: PDF icon 13_22_csmm3.pdf
  3. J. Dygaszewicz, B. Szafrański - Introducing Enterprise Architecture Framework in Statistics Poland
    Pages: 23 - 32
    Abstract: Article is devoted to the modernization of the statistical production process. The starting point for formulating the principles and methods of the modernization is to establish a framework architecture. An in-depth analysis of all relevant aspects arising from the holistic approach to the production of statistics is essential to extract the key business issues. Business needs are the foundation for the formulation of a coherent and transparent guidelines, demands, legal and technical requirements, both domestic and international which are the basis for the construction of the enterprise architecture framework and the development of solutions used to carry out the tasks of public statistics. The process of constructing architectural framework requires the use of models and methods used in scientific research, in particular arising from the achievements of enterprise architecture and modeling tools for object-oriented computing.
    Keywords enterprise architecture framework, statistics survey, GSBPM, TOGAF
    Full article: PDF icon 23_32_csmm3.pdf
  4. R. Kasprzyk, A. Stachurski - A concept of standard-based vulnerability management automation for IT systems
    Pages: 33 - 38
    Abstract: The paper focuses on the attempt to show a way of automating IT vulnerability management across enterprise systems with the use of the Security Content Automation Protocol. SCAP offers a set of components which provide, among others, adjustable security checklists, standardised dictionaries of security vulnerabilities and vulnerability scoring methods that may prove valuable for organisations in terms of security analysis activities and quantitative risk assessment.
    Keywords vulnerabilities, SCAP, security
    Full article: PDF icon 33_38_csmm3.pdf
  5. M. Zabielski, Z. Tarapata, R. Kasprzyk, K. Szkółka - Profile Cloning Detection in Online Social Networks
    Pages: 39 - 46
    Abstract: Due to the emergenceof online social networks, the importance of privacy on the Internet has vitally increased. Thus, it is important to develop mechanisms that will prevent our hidden personal data from unauthorized access and use. In this paper an attempt was made to present a concept of profile cloning detection in Online Social Networks (OSN) using Graph and Network Theory. Compering values of attributes of users’ personal profiles and analysing structural similarity of networks, we identifyattackers which steal users’ identity.
    Keywords profile cloning detection, online social networks, violations of privacy
    Full article: PDF icon 39_46_csmm3.pdf