Class imbalance remains a pressing issue in machine learning, particularly when advanced technologies require it from machine learning implementations. Medical or cybersecurity applications are good examples of this, as well as others. In this paper, we focus on cybersecurity, where the minority class often represents rare but critical attack instances. We present a comparative study of various oversampling strategies—including SMOTE-based methods, GAN-based methods, and hybrid approaches—for detecting network intrusions in imbalanced datasets. Selected classifiers are evaluated using metrics such as Accuracy, AUC, F1-score (for the minority class), and Recall. The experimental results are analysed and discussed, in terms of the strengths and weaknesses of each oversampling strategy.

Project scheduling with resource constraints plays a critical role in numerous application domains, including logistics, manufacturing, management, healthcare, and computer science. One of them is the Resource-Constrained Project Scheduling Problem with Generalized Precedence Constraints (RCPSP/max), which incorporates both minimum and maximum time lags. Due to its computational complexity, obtaining high-quality solutions for instances of even moderate size remains a difficult task. To address this challenge, a wide range of heuristic and metaheuristic approaches have been proposed in the literature. A promising direction involves parallelizing computations and combining multiple heuristic or metaheuristic methods to enhance performance and solution quality. This paper presents a set of heuristic algorithms and a parallelized, population-based multi-heuristic system implemented in the Apache Spark environment. The proposed approach offers a solution method for RCPSP/max instances and has been validated through computational experiments using benchmark datasets from the PSPLIB library.

The goal of the paper is to demonstrate that publicly available open software resources enable integration and calibration of the transformer-based classifier that performs well while solving binary classification problems. In the related work section we briefly review the most important developments in constructing transformer-based classifiers. Next, we describe components and architecture of the proposed solution, and discuss setting its hyperparameters. To validate the approach computational experiment is carried out showing that the transformer-based classifier performs well as compared with the state-of-the-art machine learning algorithms.

In this paper, we propose a uniform approach to solving a discrete-continuous scheduling problem (DCSP). DCSP is an NP-hard problem, which is a special case of the resource-constrained project scheduling problem (RCPSP). Special cases of DCSP arise in practical situations such as battery charging, minimizing energy consumption of ICT systems, or refueling a fleet of boats. To solve DCSP, its two component sub-problems must be solved: (i) sequencing tasks on machines and (ii) allocating a continuous resource to tasks in such a way as to optimize a given criterion. Problem (ii) is formulated as a convex mathematical programming problem with linear constraints that can be solved with an appropriate solver, which can be time-consuming. So far, approaches of different types have been used to solve these two sub-problems, but we propose to solve problems (i) and (ii) using the same method - differential evolution. This approach is simpler in practice than the existing ones, because it does not require inventing special methods, which makes it much easier and reduces the time to design an algorithm to solve DCSP. Computational experiment has shown that our proposed approach is highly competitive in terms of execution time to existing solutions and could be the preferred choice due to its simpler implementation.

The paper proposes an effective hybrid ACO-GA evolutionary algorithm combining the best features of the Ant Colony Optimization and Genetic algorithm. The algorithm was tested on benchmark instances of the traveling salesman problem from TSPLIB and compared to several algorithms described in the scientific literature solving the same instances. The comparison results allow us to perceive the proposed ACO-GA as an attractive compromise between the quality of solutions and the time to find a solution.

The paper proposes bicriteria oversampling strategy for mining imbalanced data. We use two specialized criteria for oversampling -classification potential and distance from the borderline between minority and majority instances. The potential is to be maximized and the distance minimized. The required number of synthetic examples is selected from the non-dominated set of examples produced by the evolutionary algorithm. At the final step the balanced set of examples is used by GEP classifier. Computational experiment confirmed that the approach assures high quality performance.

The MS-RCPSP belongs to the class of the strongly NP-hard optimisation problems. The MS-RCPSP is an extension of the classical RCPSP where some given pool of skills has been assigned to the resources. To solve this problem the multi-agent system approach has been proposed, implemented and used. The A-Team multi-agent system has been built using the environment where optimisation agents are used to find solutions. The approach has been tested experimentally using benchmark problem instances from iMOPSE dataset with the makespan optimization criterion

The paper proposes a new version of the Mushroom Picking Framework (MPF) for solving the Flexible Job Shop Scheduling problem instances. The idea is to use a set of particles, representing a set of solutions, and a special scheme for a group of solution improving agents acting in parallel. The initial set of particles is partly generated by a heuristic procedure. The approach has been validated in the computational experiment. The proposed MPF implementation produces satisfactory quality solutions in a competitive time, especially for the tasks of bigger sizes.

Abstract In the paper, we propose two gene expression programming (GEP)‐based ensemble classifiers with different drift detection mechanisms. In the related work section, we briefly review GEP as a classification tool, incremental classifiers, and concept drift detectors. Next, the structure of our two‐level GEP ensemble with metagenes is described. Further on, two integrated classifiers with drift detection algorithm and Wilcoxon rank sum test drift detector are proposed. The approach is validated in the computational experiment in which several real‐life and artificial datasets with concept drift have been used. Experiment confirmed that the proposed approach can be competitive to existing solutions. In the conclusion section, we briefly outline directions for future research.

Mining data streams require to cope with time, data size and possible concept drift constraints. Even more challenging is the case where, apart from the above, one has to deal with imbalanced data. Mining non stationary and imbalanced data streams is a relatively new area of research. In this paper, we propose the Gene Expression Programming (GEP) classifier with drift detection and data reuse for mining imbalanced data streams. GEP is used to evolve a complex expression tree returning predictions. Drift detector role is to signal the occurrence of drift which triggers inducing a new learner. Data reuse mechanism allows for improving the balance between minority and majority instances in a subset of data used for evolving the learner. The proposed approach is validated experimentally. The experiment results confirm that our classifier produces high-quality predictions.

The paper focuses on using the firefly algorithm for designing the radial-basis neural networks. The firefly algorithm belongs to a family of the global optimization tools. The firefly algorithm is a nature-inspired and one of the most powerful algorithms for solving the NP-hard optimization problems. In the paper the firefly algorithm is used as a tool for designing of the RBF network, including estimation of its output weights and transfer function parameters. The details of the implementation are discussed. Computational experiment has been carried-out with a view to investigate effectiveness of the discussed implementation. Experiment results have confirmed usefulness of the proposed approach. Conclusions include suggestions for future research.

The paper focuses on using stacking and rotation-based technique to improve performance and generalization ability of the machine learning classification with data reduction. The aim of data reduction technique is decreasing the quantity of information required to learn a high quality classifiers, especially when the data are huge. The paper shows that merging both stacking and rotation-based ensemble techniques with machine classification based on data reduction may bring additional benefits with respect to the accuracy of the classification process. The finding that has been confirmed by computational experiments. The paper includes the description of the approach and the discussion of the computational experiment results.

This volume comprises proceedings of the 2017 IEEE International Conference on INnovations in Intelligent SysTems and Applications (INISTA 2017), organized by Gdynia Maritime University, Poland, in cooperation with Yildiz Technical University, Turkey, IEEE Systems, Man, and Cybernetics Society (SMC), IEEE Poland Section, Poland Section of IEEE Systems, Man, and Cybernetics Society Chapter, IEEE SMC Technical Committee on Computational Collective Intelligence and Gdynia Maritime University's Students and Alumni Foundation. Technical Support of the Conference has been provided by the Poland Section of IEEE Computer Society Chapter. The INISTA 2017 took place in Gdynia, Poland, on July 3–5, 2017.

This paper focuses on the machine classification with data reduction. The aim of the data reduction techniques is decreasing the quantity of information required to learn a high quality classifiers. In this paper the data reduction is carried out by selection of relevant instances, called prototypes. To solve the machine classification problem with data reduction an agent-based population learning algorithm is proposed. The discussed approach bases on the assumption that the selection of prototypes is carried-out by a team of agents and that the prototype instances are selected from clusters of instances. The proposed procedure is called the stack generalization. It aims at improving the quality of the learning process and increasing the generalization capacity of the learning model. The paper includes the description of the approach and the discussion of the validating experiment results.

The article addresses the problem of feature selection in a kernel space and proposes the approach to selecting the most informative features for classification carried out using the RBF neural classifier. In the article, class-dependent and cluster-dependent methods for feature selection are considered. The process of feature selection is supported by the rotation-based ensemble technique. The feature selection problem is viewed as an optimization task solved by the agent-based population learning algorithm applied at the RBFN’s initialization and training stage. The proposed approach is validated experimentally, and the obtained results are compared with the results produced using other methods. Experiment results show that the proposed method of feature selection in a kernel space of the RBF neural networks can be considered as a useful approach to constructing high-quality RBFN-based classifiers.

The problem addressed in this study concerns mining data streams with concept drift. The goal of the article is to propose and validate a new approach to mining data streams with concept-drift using the ensemble classifier constructed from the one-class base classifiers. It is assumed that base classifiers of the proposed ensemble are induced from incoming chunks of the data stream. Each chunk consists of prototypes and information about whether the class prediction of these instances, carried-out at earlier steps, has been correct. Each data chunk can be updated by using the instance selection technique when new data arrive. When a new data chunk is formed, the ensemble model is also updated on the basis of weights assigned to each one-class classifier. In this article, two well-known instance-based learning algorithms—the CNN and the ENN—have been adopted to solve the one-class classification problems and, consequently, update the proposed classifier ensemble. The proposed approaches have been validated experimentally, and the computational experiment results are shown and discussed. The experiment results prove that the proposed approach using the ensemble classifier constructed from the one-class base classifiers with instance selection for chunk updating can outperform well-known approaches for data streams with concept drift.

The problem addressed in this paper concerns mining data streams with concept drift. The goal of the paper is to propose and validate a new approach to mining data streams with concept-drift using the ensemble classifier constructed from the one-class base classifiers. It is assumed that base classifiers of the proposed ensemble are induced from incoming chunks of the data stream. Each chunk consists of prototypes and can be updated using instance selection technique when a new data have arrived. When a new data chunk is formed, ensemble model is also updated on the basis of weights assigned to each one-class classifier. The proposed approach is validated experimentally.

In the paper, we propose an island-based differential evolution algorithm (IBDEA) for solving the discrete-continuous scheduling problem (DCSP) with continuous resource discretisation - ΘZ. The considered problem originates from DCSP, in which nonpreemtable tasks should be scheduled on parallel identical machines under constraint on discrete resource and requiring, additionally, a renewable continuous resource to minimize the schedule length. The continuous resource in DCSP is divisible continuously and is allocated to tasks from a given interval in amounts unknown in advance. Task processing rate depends on the allocated amount of the continuous resource. To eliminate time consuming optimal continuous resource allocation, an NP-hard problem ΘZ with continuous resource discretisation is introduced and sub-optimally solved by IBDEA. Experimental results show that IBDEA is able to find better solutions than an algorithm realizing only the differential evolution method and was able to improve best-known solutions to the considered problem.

This article proposes a novel approach to the radial basis function network (RBFN) design. Its main idea is to apply the agent-based population learning algorithm to the task of initialization and training RBFNs. The approach allows for an effective network initialization and estimation of its output weights. The initialization involves two stages, where in the first one initial clusters are produced using the similarity-based procedure and next, in the second stage, prototypes (centroids) from the thus-obtained clusters are selected. The agent-based population learning algorithm is used to select prototypes. In the proposed implementation of the algorithm, both tasks—RBFN initialization and RBFN training—are carried out by a team of agents executing various local search procedures and cooperating with a view to determine the solution to the RBFN design problem at hand. The performance of the RBFN constructed using the proposed agent-based approach is analyzed and evaluated. The proposed approach is also compared with different RBFN initialization and training procedures in the literature.

Clustering techniques have an influence on the quality and performance of RBF networks. The aim of the paper is to propose and evaluate a similarity-based approach for cluster initialization with the agent-based population learning algorithm used to select prototypes from the obtained clusters. The performance of the RBF network constructed using the proposed clustering and selection approach is analyzed and evaluated. The proposed approach is also compared with different RBF network initialization procedures.

In this paper we propose a simple and effective combined classifier based on the data reduction carried-out through applying fuzzy C-means clustering and differential evolution techniques. The idea is to produce clusters from the training set instances applying fuzzy C-means algorithm. In further step cluster centroids are used as seeds in the differential evolution algorithm to construct prototypes, each representing a single cluster. Simple distance-based weak classifiers are then used to produce the Ada Boost combined classifier. The approach has been validated experimentally. Computational experiment results confirm good quality of the proposed classifier.

In this paper the Team of A-Teams (TA-Teams) architecture for solving the resource-constrained project scheduling problem (RCPSP) is proposed and experimentally validated. RCPSP belongs to the NP-hard problem class. To solve this problem a parallel cooperating A-Teams consisting of an asynchronous agents implemented using JABAT middleware have been proposed. From one to four A-Teams and four kinds of optimization agent have been used. Computational experiment involves evaluation of the proposed approach in respect of the different parameters settings controlling the working and migration strategies used in the Team of A-Teams approach.

The paper sums up the impact of two parameters defining the migration process, namely topology and frequency, in a systems with A-Teams working in parallel, in the architecture designed for solving difficult combinatorial optimization problems. A-Teams, belonging to the team of ATeams, cooperate through exchange of intermediary computation results. The process of forwarding a result from one A-Team to another is called the migration. Several known migration models have been compared. We propose and test also an original model of communication, called Randomized, with no predefined migration topology. In this case migrations take place not with a predefined frequency, but only after an A-Team has not improved its best current solution for some given time. The proposed migration model outperforms all the remaining ones under investigation.

The aim of this article is to propose and evaluate an agent-based population learning algorithm generating, through prototype selection, a representative training data set of the required size. The main feature of the proposed approach is selection of the representative feature vectors, called prototypes, from clusters of feature vectors constructed over the original training data set under the assumption that from each cluster a single prototype is obtained. Thus, the number of clusters produced from the original training data set has a direct influence on the size of the reduced data set. The process of selection is executed by a team of agents, which execute various local search procedures and cooperate to determine a solution to the instance reduction problem, aiming at obtaining a compact representation of the data set. Rules for agent cooperation during the clustering and selection processes are defined within the so-called working strategy used by the team of agents (A-Team) in question. The article proposes a set of procedures that are used by agents to produce clusters and select prototypes. The approach is validated experimentally using well-known benchmark data sets. In addition to the computational experiment used to validate the model, the article investigates the efficiency and performance of two different working strategies used by the proposed A-Team. Because the proposed approach is based on the population learning algorithm, which belongs to the class of the population-based methods, an evaluation of the influence of the population of solution size on the performance of the algorithm is also included.

The paper deals with the distributed learning. Distributed learning from data is considered to be an important challenge faced by researchers and practice in the domain of the distributed data mining and distributed knowledge discovery from databases. An effective approach to learning from a geographically distributed data is to select, from the local databases, relevant local patterns, called also prototypes. Such a selection can be based on results of the data reduction process. The paper proposes to carry-out prototype selection at local sites in parallel, independently at each site, employing specialized software agents. To assure obtaining homogenous prototypes at a global level the consensus-based method is proposed and applied. The paper includes a detailed description of the proposed approach and a discussion of the computational experiment results.

The purpose of this article is to present the application of the A-team approach to solving some machine learning problems belonging to the supervised and unsupervised learning classes. Because the above problems are computationally hard, it is proposed to take advantage of the robustness and flexibility of population-based methods combined with the efficiency of multi-agent systems integrated within the A-team concept. The main part of the article summarizes the experiences of the authors gained while developing various A-teams and includes some examples of population-based multi-agent algorithms for solving problems from the machine learning domain. It can be concluded that population-based multi-agent algorithms can be competitive in comparison with other existing techniques for some machine learning problems.

Today, a significant number of the approaches that enhance the performance of real-life systems are based on knowledge processing intelligent multi-agent methodologies. This is a new paradigm that ...

The paper compares a new population based method called PLA (population learning algorithm) with the evolutionary computation approach. The paper introduces a concept of the population learning algorithm, compares its features with those of the evolutionary computation approach and presents the results of the computational experiment. It has involved using both – PLA and evolutionary computation to solving a randomly generated set of instances of several computationally difficult combinatorial optimization problems. Experiment result show that using PLA to solve some difficult problems can produce competitive results both in terms of the computational effort required and quality of solutions.

The systematic approaches to testing and reliability determination of programs e.g. are applicable to imperative programming but not immediately to declarative programming, such as logic programming, which is of great importance to develop knowledge-based systems. We describe an approach to implementation-based testing and reliability determination of logic programs materialized in a product assurance environment, presently limited to two major components: the test environment PROTest and the reliability assessment environment PRORool, with the results of the former serving as input for the latter. The test environment consists of structure analysis of logic programs, automatic test case generation and execution, test coverage determination, and generation of test reports. The reliability assessment environment provides an approach to reliability prediction and estimation of Prolog programs, introducing two measures describing Prolog programs complexity, which are used to determine the program reliability. It implements also several well-known software reliability models for comparison purposes.

Consensus recovery block scheme and concurrent recovery scheme may become useful for real-time applications as they consider concurrency and time efficiency. The authors compare the performance of both techniques in terms of the times they consume to achieve software fault tolerance. For this purpose a simple model is used and analysed which simulates different real-time situations.>

An approach to the optimization of software reliability is proposed. The emphasis is put on the software redundancy to achieve fault tolerance, i.e. the results of the optimization process are used to determine the optimal structure of the software to be developed. Two optimization models are formulated covering, respectively, modified recovery block scheme and multiversion programming approaches. Both cases are illustrated by simple examples. The models show that it is possible to formulate and solve some software related reliability optimization problems. They further show that the concept of redundancy to achieve fault tolerance (basic for the traditional theory of reliability) can be used in the field of software reliability optimization.>