This paper presents a novel method of text categorization based on the use of specialized dictionaries.The method is applied to texts of mass media and short comments on the Internet. The analysis indicates the effectiveness and efficiency of word stemming for text categorization problem and the validity of the proposed method
The article considers the process of building an intrusion detection system using intelligent network traffic analysis. The requirements for the developed system of intrusion detection are formulated, as well as its architecture is proposed. As a mechanism for making decisions about the presence of attacks, it is suggested to use methods of inductive machine learning, namely, artificial neural networks. The paper proposes the construction of a neural network model based on a multilayer perceptron, for which the most significant input parameters are determined. The technique of constructing the intelligent network traffic analysis module, its logic of work are considered. The client-server application for network traffic analysis on the generated parameters was developed ang the results of testing are given in the paper. The created module of intelligent network traffic analysis shows high accuracy of attacks identification. To increase the accuracy of network attack classification, in future studies, it is planned to supplement the intelligent network traffic analysis module with other methods of machine learning, in particular, the machine classifier.
The paper proposes a method of multicriteria optimization under interval stochastic uncertainty of estimates given by the subject for the relative importance of one criterion over the other and the different alternatives to each other for each criterion. The method is an extension of the deterministic Analytic Hierarchy Process (AHP) for multicriteria optimization. It is use deterministic point estimates of the importance of criteria and alternatives for each criterion . While deterministic AHP allows to select the best alternative by a point maximum value of a global priority in the developed article interval stochastic AHP the global priorities are interval, making it difficult to make the best decision . To select the best interval alternative in this article introduce two criteria, whose values are maximized. The first criterion corresponds to the maximum of the lower and upper bounds of the intervals of global priorities of alternatives. The second criteria is the maximum of interval stability of alternatives. Application of the proposed approach is illustrated by a specific example. Also a comparison with the results obtained on the basis of interval arithmetic, show the failure of the latter, carried out.
In clinical trials comparing experimental and control treatment the effect of treatment often depends on the range of patient’s characteristics (biomarkers) such as clinical, anthropological, genetic, psychological, social characteristics and others. Personalized medicine aims at finding such dependencies to tailor treatment strategies to a patient. This paper presents an overview of the approaches to data analysis of clinical trials intended for identification of influential biomarkers and subgroups of patients, where experimental and control treatment differ significantly in efficiency.
In work a number of methodological questions of evolutionary calculations is critically considered. The conclusion that within existing representations the AI evolutionary direction hardly can be beyond methods of search optimization is drawn. The number of questions which decision can turn evolutionary calculations into really main direction of development of intellectual systems is offered.