Ninoslav Slavek

Improved Bisector pruning for uncertain data mining

Uncertain data mining is well studied and very challenging task. This paper is concentrated on clustering uncertain objects with location uncertainty. Uncertain locations are described by probability density function (PDF). Number of uncertain objects can be very large and obtaining quality result within reasonable time is a challenging task. Basic clustering method is UK-means, in which all expected distances (ED) from objects to clusters are calculated. Thus UK-means is inefficient. To avoid ED calculations various pruning methods are proposed. The pruning methods are significantly more effective than UK-means method. In this paper, Improved Bisector pruning method is proposed as an improvement of clustering process.

The concept of quality information system (QIS)

The product quality characteristics should be the prime drivers when assessing and improving the quality of the software development process as we are concerned with the product quality. The quality of the software product is determined by the quality of the software process. This seems intuitive but there is no empirical evidence to prove its validity yet. QIS establishes a system that enables to analyse the relation between base practices and processes of the SPICE model and the eleven product quality factors and criteria of McCalls (1977) model for software product evaluation. The main goal of QIS is to evaluate and verify benefits gained by improving the process maturity level. In front line of both, the process model and product quality model is the software product improvement, resulting in a high quality software product delivered on time and at less cost.

Positioning of Public Service Systems Using Uncertain Data Clustering

Positioning of public service system is crucial and very challenging task. Proper positioning ensures that the public service system would complete its tasks to the end users. This paper is focused on finding the best location for public service system, to improve its efficiency when using uncertain data clustering. By choosing the best location for the service system the respond time can be minimised, and the given tasks could be performed in a reasonable time. Improved bisector pruning method was proposed for clustering previous data of public service system to find the best location for its application. Presented method can be used for different Public Service Systems, like traffic services, positioning of ambulance vehicles and other mobile objects. Cluster centres are used as best locations for public service systems because, cluster centres minimized total expected distance from tasks that have been set to the service system. On this way, public service system will be improved and can fulfil more tasks during the shortest period of time.

Computer Simulation of Differential Digital Holography

This paper presents a complete procedure of the differential digital holography process. The method is based on digital hologram construction and reconstruction. Holograms and reconstructs are calculated by applying Fourier Transformation on bitmap images. The elaborated algorithm allows amplification of small differences between two images of the same size. Advantages and disadvantages of the described process are discussed.

Grey System Theory Approach to Quality of Intranet

This work discussed the effectiveness of he Gray system theory to determine the Intranet quality. ISO 9126 [1] is used to define the Intranet quality factors and characteristics. The ranking of these factors and characteristics were used to determine the importance of this quality attributes. The Grey system provides multidisciplinary approaches for analysis and abstract modelling of systems for which the information is limited, incomplete and characterized by random uncertainty. The Grey relational analysis uses information from the Grey system to dynamically compare each characteristic quantitatively.

REINFORCEMENT LEARNING IN NON-MARKOV CONSERVATIVE ENVIRONMENT USING AN INDUCTIVE QUALITATIVE MODEL

The majority of real-world processes, such as power plants, banking and retail businesses, are non-Markov processes, being conservative systems with stochastic supply and demand. As an example, a retail process possesses long-term memory of the customers experience and market price drift that deviates from the Markov property. Modeling the reward in this process is directed towards actions that have to be executed daily in order to support it. These actions are further severely distracted by the hidden periodicity of customer behavior on a monthly and weekly basis. Alternative solutions in the retail business are achieved using a retail potential market model and a pricing policy based on demography. The policy of non-Markov behavior has not been intensively studied, although the literature indicates the non-Markov nature of many real process models, such as bank rating migrations. A solution is proposed, based on day-to-day data collection from point-of-sale (POS) locations, synthesizing the reward function from separate sale component rewards using qualitative models, and indicating the most outstanding sale groups that form the reward model. The normalization of POS data has been used for the elimination of periodicities and of non-Markov features of the process data. Reinforcement learning has been additionally supported by artificial corrections of the normalized reward function, and thus the obtained models used for recognition of the most promising and most defective hidden retail product groups. Model data were analyzed for the statistical significance of the obtained results, comparing normalized and non-normalized sales data distributions. The method is simple and effective, being applicable to each POS separately, for a complex retail business network, as well as for other conservative environments. The obtained qualitative correlations of model and reward function lie between 0.72 and 0.95, even for the simple cases presented.