Event-Based Life in a Nutshell: How Evaluation of Individual Life Cycles Can Reveal Statistical Inferences Using Action-Accumulating P Systems

摘要

A sequence of perceivable events or recorded observations over time commonly witnesses the life cycle of an individual at a macroscopic perspective. In case of a human being, birth could make the starting point followed by successive maturation along with increase of individual skills. Further events like foundation of a family, stages of career, coping with dramatic diseases, loss of abilities, and finally the death mark crucial events within a human life cycle. Even beyond biology, life cycles are present in various contexts, for instance when elucidating the quality of durable technical products such as cars. Social scenarios or games with several players incorporate consideration of life cycles as well. Provided by logfiles or monitoring reports, dedicated accumulation of events facilitates identification of life cycles whose statistical analysis promises valuable insights. To this end, we formalise an individual by a set of attributes. Based on its underlying initial assignment ("genetic potential"), events can update corresponding attribute values. Furthermore, events might create new individuals but also kill or merge existing ones. For modelling and evaluation of life cycles, we introduce action-accumulating P systems inspired by dealing with events which in turn result in actions at the system's level. Two case studies demonstrate practical benefits from our approach: We explore the survival of pieces in the board game Mensch argere Dich nicht (Man, don't get annoyed - a variation of Ludo). Secondly, we interpret pseudonymised data from 1,108 students who attended our university course Introduction to Programming stating main factors to improve the final grade with emphasis on the effect of passing a line of exercises and practical training offers.