P colony is a formal computational model suitable for modelling behaviour of simple agents acting in a shared environment. It builds on the framework of membrane systems, where both the environment and agents contain abstract discrete objects and formal rules acting upon them. Adding a 2D geometrical structure and an evolution of the environment results in the model of 2D evolving P colonies. The model is suitable for the simulation of phenomena like stigmergy, hence also for the implementation of multi-agent optimisation strategies. The main motivation lies in the use of P colonies as a high-level description language for optimisation tasks, reconfigurable without programming, and ready for a highly parallel and efficient implementation by existing GPU libraries for P systems. In this paper, we use a 2D P colony to implement an ant colony-inspired optimisation algorithm. The agents—ants—search the environment for food representing the extrema of an objective function. The search is oriented with the help of pheromone trails left by previous agents. The trails are subject to a decay and can eventually vanish. In our previous research, we implemented via a 2D P colony a classical formulation of the ant algorithm where ants immediately collect found food in the nest. Here we introduce an improved 2D P colony implementation inspired by variants of ACO algorithms with local search. We call this variant a vicinity search and we demonstrate experimentally that it greatly improves the search results. Furthermore, the new solution decreases asymptotically the complexity of programs in agents.
To obtain more results in the field of theoretical models of the multiagent systems it is necessary to step outside of the research done on the paper and move forward to the computer simulation. The simulation opens the possibilities to study the behavior of studied model in real-time, or verify designed configurations. To develop an application of a good quality and fulfilling all the needs, the development of such simulator must pass standard phases of the information system life cycle. In this paper, we present a development of a simulator of a 2D P colony with the evolving environment simulator. The simulator replaces the previous version of the software we used to simulate the behavior of the 2D P colony with evolving environment, where the configuration of the colony was hardcoded into the source code of the simulator, hence, any changes had to be done by developing a new version of the simulator. Recent version is modular and the definition of the simulated colony is held in the separate configuration file. During the development of the application, we followed phases of the information system life cycle. The result of the development is an application in the Python programming language with desired features. The application was developed under the Open Source license; hence, it is freely accessible and can be used or modified by anyone. MSC Classification: 68Q01 , 68Q07 , 68Q42
3 March 2026 in International Journal of Computational Intelligence Systems
The cardiotocogram (CTG) is a technique used to evaluate the health of the baby during pregnancy. Uterine contractions (UC) and Fetal Heart Rate (FHR) are the two main outcomes of CTG. Twenty-one parameters, including baseline values, accelerations, foetal movements, light decelerations, etc., are included in the assessment of FHR and UC on CTG. Obstetricians can use these parameters to help them assess if the fetus’s health is normal, suspicious, or abnormal. A variety of machine learning methods can be used to assess the health of the fetus. This study looks at the K-Nearest Neighbor (KNN), Decision Tree, Random Forest (RF), Gradient Boosting and proposed model PCBoost (Principal Component Analysis-Gradient Boosting) for predicting fetal hypoxia. This study also compares the performance metrics (precision, recall, F1 score, accuracy and kappa score) of all the above algorithms. The proposed model uses Principal Component Analysis (PCA) implementation for dataset size reduction to assess the performance. The study produces positive findings, as the proposed PCBoost model achieves a maximum accuracy of 99% under Scenario S4, where PCA retains 97% variance using 16 components, outperforming other machine learning techniques. While the current study relies on the UCI CTG dataset for evaluation, efforts were made to prevent overfitting through PCA and cross-validation techniques. Future work will include validation on additional clinical datasets to enhance model robustness and clinical applicability. This illustrates how well the proposed models enhance the precision of the CTG-based classification of fetal health. The results support the expediting of fetal health evaluations by incorporating machine learning models into standard clinical procedures. The study emphasizes the need of early issue detection while demonstrating how machine learning (ML) may be utilized to maximize medical resource allocation and time efficiency.
To obtain more results in the field of formal models of multi-agent systems it is necessary to step outside of the research done on the paper and move forward to computer simulation. The simulation opens the possibilities to study the behavior of studied model in real-time, or verify designed configurations. To develop suitable application of a good quality and fulfilling all the needs, it is necessary to base the development on detailed analysis. In this paper, we present a development of a simulator of a 2D P colony with the evolving environment. The simulator replaces the previous version of the software we used to simulate the behavior of the 2D P colony with evolving environment. In the previous version, the configuration of the colony was hardcoded into the source code of the simulator; hence, any changes had to be done by developing a new version of the simulator. Recent version is modular and the definition of the simulated colony is held in the separate configuration file. The result of the development is an application in the Python programming language with desired features defined in the analysis of the application. The application was developed under the Open Source license; hence, it is freely accessible and can be used or modified by anyone.