Bartosz A. Nowak

Multi-class Nearest Neighbour Classifier for Incomplete Data Handling

The basic nearest neighbour algorithm has been designed to work with complete data vectors. Moreover, it is assumed that each reference sample as well as classified sample belong to one and the only one class. In the paper this restriction has been dismissed. Through incorporation of certain elements of rough set and fuzzy set theories into k-nn classifier we obtain a sample based classifier with new features. In processing incomplete data, the proposed classifier gives answer in the form of rough set, i.e. indicated lower or upper approximation of one or more classes. The basic nearest neighbour algorithm has been designed to work with complete data vectors and assumed that each reference sample as well as classified sample belongs to one and the only one class. Indication of more than one class is a result of incomplete data processing as well as final reduction operation.

An Application of Firefly Algorithm to Position Traffic in NoSQL Database Systems

In this paper, an application of Computational Intelligence methods in positioning and optimization of traffic in NoSQL database system modeled with exponentially distributed service and vacation is discussed. Positioning of the system modeled with independent 2-order hyper exponential input stream of packets and exponential service time distribution is solved using firefly algorithm. Different scenarios of examined system operation are presented and analyzed.

The Learning of Neuro-Fuzzy Classifier with Fuzzy Rough Sets for Imprecise Datasets

The paper concerns the architecture of a neuro-fuzzy classifier with fuzzy rough sets which has been developed to process imprecise data. A raw output of such system is an interval which has to be interpreted in terms of classification afterwards. To obtain a credible answer, the interval should be as narrow as possible; however, its width cannot be zero as long as input values are imprecise. In the paper, we discuss the determination of classifier parameters using the standard gradient learning technique. The effectiveness of the proposed method is confirmed by several simulation experiments.

Application of Rough Sets in k Nearest Neighbours Algorithm for Classification of Incomplete Samples

Algorithm k-nn is often used for classification, but distance measures used in this algorithm are usually designed to work with real and known data. In real application the input values are imperfect—imprecise, uncertain and even missing. In the most applications, the last issue is solved using marginalization or imputation. These methods unfortunately have many drawbacks. Choice of specific imputation has big impact on classifier answer. On the other hand, marginalization can cause that even a large part of possessed data may be ignored. Therefore, in the paper a new algorithm is proposed. It is designed for work with interval type of input data and in case of lacks in the sample analyses whole domain of possible values for corresponding attributes. Proposed system generalize k-nn algorithm and gives rough-specific answer, which states if the test sample may or must belong to the certain set of classes. The important feature of the proposed system is, that it reduces the set of the possible classes and specifies the set of certain classes in the way of filling the missing values by set of possible values.

The learning of neuro-fuzzy approximator with fuzzy rough sets in case of missing features

The architecture of neuro-fuzzy systems with fuzzy rough sets originally has been developed to process with imprecise data. In this paper, the adaptation of those systems to the missing features case is presented. However, the main considerations concern with methods of learning which could be applied to such systems for approximation tasks. Various methods for determining values of system parameters have been considered, in particular the gradient learning method. The effectiveness of proposed methods has been confirmed by many simulation experiments, which results have been supplied to this paper.

A New Method of Improving Classification Accuracy of Decision Tree in Case of Incomplete Samples

In the paper a new method is proposed which improves the classification accuracy of decision trees for samples with missing values. This aim was achieved by adding new nodes to the decision tree. The proposed procedure applies structures and functions of well-known C4.5 algorithm. However, it can be easily adapted to other methods, for forming decision trees. The efficiency of the new algorithm has been confirmed by tests using eleven databases from UCI Repository. The research has been concerned classification but the method is not limited to classification tasks.

Genetic fuzzy classifier with fuzzy rough sets for imprecise data

The main problem addressed in this paper is to handle adequately imprecision of input data by means of a combination of fuzzy methods with the rough set theory. We will make use of fuzzy rough sets derived as rough approximations of fuzzy antecedent sets by non-singleton fuzzy premise sets in a fuzzy classifier. Adaptation of the parameters of this system will be done by the standard genetic algorithm.