Wyniki wyszukiwania - BazTech

1

Application of random sampling in the concept-dependent granulation method

Cybulski Radosław, Artiemjew Piotr

Annals of Computer Science and Information Systems

|

2022

|

Vol. 31

3--11

EN

Professor Zadeh in his works proposed the idea of grouping similar objects on the basis of certain similarity measures, thus initiating the paradigm of granular computing. He made the assumption that similar objects may have similar decisions. This natural assumption, operates in other scientific methodologies, e.g. methods based on k nearest neighbours, in reasoning by analogy and in rough set theory. The above assumption implies the existence of grouped information nodes (granules) and has potential applications in reducing the size of decision systems. The hypothesis has guided, among others, the creation of granulation techniques based on the use of rough inclusions (introduced by Polkowski and Skowron) - according to the scheme proposed by Polkowski. Where the possibility of a large reduction of the size of decision systems while maintaining the classification efficiency was verified in experimental works.In this paper, we investigate the possibility of using random sampling in the approximation of decision systems - as part of dealing with big data sets. We use concept-dependent granulation as a reference approximation method. Experiments on selected real-world data have shown a common regularity that gives a hint on how to apply random sampling for fast and effective size reduction of decision systems.

2

About Classifiers Quality Assessment: Balanced Accuracy Curve (BAC) as an alternative for ROC and PR Curve

Weiss Aleksandra, Młyński Marcin, Artiemjew Piotr

Annals of Computer Science and Information Systems

|

2022

|

Vol. 30

149--156

EN

In this work, we propose a new parameter to study the effectiveness of classifiers - the AUC (area under curve) of the balanced accuracy curve (BAC) on data with different balance degrees - we compare its effectiveness with the popular AUC parameters for the ROC and PR curve. We use a global kNN classifier with typical metrics to verify the utility of the new parameter. BAC, ROC and PR curves generate similar results, the advantage of BAC is its simplicity of implementation and ease of interpretation of results.

3

A Novel Ensemble Model - The Random Granular Reflections

Artiemjew Piotr, Ropiak Krzysztof

Fundamenta Informaticae

|

2021

|

Vol. 179, nr 2

183--203

EN

One of the most popular families of techniques to boost classification are Ensemble methods. Random Forests, Bagging and Boosting are the most popular and widely used ones. This article presents a novel Ensemble Model, named Random Granular Reflections. The algorithm used in this new approach creates an ensemble of homogeneous granular decision systems. The first step of the learning process is to take the training system and cover it with random homogeneous granules (groups of objects from the same decision class that are as little indiscernible from each other as possible). Next, granular reflection is created, which is finally used in the classification process. Results obtained by our initial experiments show that this approach is promising and comparable with other tested methods. The main advantage of our new method is that it is not necessary to search for optimal parameters while looking for granular reflections in the subsequent iterations of our ensemble model.