Marina V. Fomina

Problem of knowledge discovery in noisy databases

The problem of information generalization for real data that may contain noisy data is considered. Various models of information noise are presented, and the influence of noise to the algorithms of generalization is discussed. We used the methods of constructing decision trees and forming production rules. The results of the modeling are presented.

Argumentation Approach and Learning Methods in Intelligent Decision Support Systems in the Presence of Inconsistent Data

Abstract This paper contains a description of methods and algorithms for working with inconsistent data in intelligent decision support systems. An argumentation approach and application of rough sets for generalization problems are considered. The methods for finding the conflicts and the generalization algorithm based on rough sets are proposed. Noise models in the generalization algorithm are viewed. Experimental results are introduced. A decision of some problems that are not solvable in classical logics is given.

Methods and algorithms of information generalization in noisy databases

The problem of information generalization with account for the necessity of processing the information stored in real data arrays which may contain noise is considered. Noise models are presented, and a noise effect on the operation of generalization algorithms using the methods of building decision trees and forming production rules is developed. The results of program modeling are brought about.

Methods and Algorithms of Anomaly Searching in Collections of Time Series

One of the basic problems arising under processing temporal dependencies is the analysis of time series. The various approaches to processing of temporal data are considered. The problem of anomaly detection among sets of time series is setting up. The algorithm TS-ADEEP-Multi for anomaly detection in time series sets for the case when the learning set contains examples of several classes is proposed. The method for improving the accuracy of anomaly detection, due to “compression” of these time series is used. Modelling results for anomaly detection in time series are produced.

Inductive inference and argumentation methods in modern intelligent decision support systems

Improvement of the classification quality for the generalization problem is considered. In order to improve the classification models produced by generalization algorithms, it is proposed to use argumentation methods based on defeasible reasoning with justification degrees. New methods and algorithms are proposed, and experimental results on various test data sets are described, including instances with noisy initial data.

Protection of Information in Networks Based on Methods of Machine Learning

The paper considers the possibility of using artificial intelligence methods in information security tasks: methods for generating inductive concepts for analyzing network traffic, as well as methods of argumentation for automated security decision support systems. The approach proposed in the work allows giving quantitative assessments of the quality of the recommendations developed by the system, thereby helping to solve an important task - the task of choosing the way of responding to suspicious activity in the system. Examples of handling dangerous situations occurring in the system are also presented.

Application of Intelligent Data Analysis Methods for Information Security Problems

In this paper, we consider the ideas and approaches to the intelligent data analysis methods in problems of information security. Solving network security problems is a complex task, involving the large number of factors and requiring finding reasonable compromises between maintaining security, the stable work, enhancing operating expenses and functional restrictions of complex information systems. There are considered the ways to apply inductive concept formation methods for analyzing network traffic, as well as argumentation methods for an automated support of security solutions. The proposed approach allows to give numerical assessments of the quality of recommendations developed by the system, thereby helping to solve an important task - the task of choosing the way to react to suspicious activity in the system. In addition, the example of handling the dangerous situations arising in the system is given.

Argumentation in inductive concept formation

This paper contains a description of methods and algorithms for solving the generalization problem in intelligent decision support systems. For this purpose the argumentation approach for inductive concept formation is used. The methods for finding the conflicts and the generalization algorithm based on the rough set theory are proposed. It is suggested to use the argumentation, based on defeasible reasoning with justification degrees, to improve the quality of the classification models obtained by the generalization algorithm. Noise models in the generalization algorithm are viewed. Experimental results are introduced.

Modeling of algorithms of inductive concept formation in noisy databases

The problem is inductive concept formation in the case of the processing of incomplete, inaccurate, and inconsistent information stored in real data sets. In order to generalize information from real databases it is proposed to use production models and decision trees. Models of noise are presented and the effect of noise on the operation of the proposed generalization algorithms is examined. The results of the program modeling are given.