Arwin Datumaya Wahyudi Sumari, Rosa Andrie Asmara, Dimas Rossiawan Hendra Putra, Ika Noer Syamsiana
Knowledge Growing System (KGS) is a new perspective in Artificial Intelligence (AI) that is belonged to Cognitive AI (CAI). It is built based on the constructivism theory of the cognitive psychology field. KGS learning ability emulates how human learns through interactions over time with a phenomenon. This new CAI technology has been implemented in various use-cases such as weather prediction, face recognition, dynamics prediction, strategic decision making. The implementation is also not limited to enhancing the security of electronics hardware, energy usage prediction, flood early warning system, and translated into a cognitive processor. In its basic design, KGS uses binary-type extracted features as the inputs because it was designed for firm decision-making support. In this paper, we propose using KGS for non-binary-type decision support systems and observe its performance in pattern recognition use-case. The recognition was done to the Iris dataset. The result shows that KGS has recognized the Iris with an accuracy of up to 90.91%. This result shows that KGS is prospective for non-binary-type input decision support system. © 2023 Author(s).
Cognitive Artificial Intelligence Research Group (CAIRG), Indonesia; Department of Electrical Engineering, Politeknik Negeri Malang, Jl. Soekarno-Hatta 9, Jawa Timur, Malang, 65141, Indonesia; Department of Information Technology, Politeknik Negeri Malang, Jl. Soekarno-Hatta 9, Jawa Timur, Malang, 65141, Indonesia; Abdulrachman Saleh Air Force Base, 2nd Operation Command, Indonesian Air Force Malang, Jawa Timur, 65154, Indonesia